Geology and Hydrogeochemistry of the Jungapeo CO2-Rich Thermal

Journal of Volcanology and Geothermal Research 163 (2007) 1–33 www.elsevier.com/locate/jvolgeores

Geology and hydrogeochemistry of the Jungapeo CO2-rich thermal springs, State of Michoacán, Mexico ⁎ Claus Siebe a, , Fraser Goff b, María Aurora Armienta a, Dale Counce c, Robert Poreda d, Steve Chipera c

a Departamento de Vulcanología, Instituto de Geofísica, Universidad Nacional Autónoma, de México, Coyoacán 04510, México D.F., México b Department of Earth and Planetary Sciences, University of New Mexico, Albuquerque, NM 87131, USA c Earth and Environmental Sciences Division, Los Alamos National Lab., Los Alamos, NM 87544, USA d Department of Earth and Environmental Sciences, University of Rochester, Rochester, NY 14627, USA

Received 19 April 2006; received in revised form 28 March 2007; accepted 30 March 2007 Available online 20 April 2007

Abstract

We present the first geothermal assessment of the Jungapeo CO2-rich mineral springs, which are located in the eastern part of Michoacán State (central Mexico) at the southern limit of the Trans-Mexican Volcanic Belt. All but one of the N10 springs occur at the lower contact of the distal olivine-bearing basaltic andesite lavas of the Tuxpan shield, a 0.49- to 0.60-Ma-old cluster of monogenetic scoria cones and lava flows. The Tuxpan shield has a maximum radius of 6 km and was constructed on top of a folded and faulted Cretaceous basement consisting largely of marine limestones, marls, and shales. The mineral waters are characterized by moderate temperatures (28 to 32 °C), mild acidity (pH from 5.5 to 6.5), relatively high discharge rates, effervescence of CO2 gas, clarity at emergence and abundant subsequent precipitation of hydrous iron, silica oxides, and carbonates around pool margins and issuing streamlets. Chemical and isotopic (deuterium, oxygen, and tritium) analyses of water and gas samples obtained during the period 1991–1997 indicate that the springs are largely composed of meteoric water from a local source with relatively short residence times (water ages of 7 to 25 years). Spring waters are chemically characterized by moderate SiO2, Ca+Mg nearly equal to Na+K, high HCO3, moderate to low Cl, low F and SO4, high B, moderate Li, while Br and As are low. In contrast, Fe+Mn is exceptionally high. Thus, the Jungapeo waters cannot be regarded as high-temperature geothermal fluids. Instead, they resemble soda spring waters similar to other low-to-medium temperature soda waters in the world. Gas samples are extremely rich in CO2 with no detectable geothermal H2SorH2 and very low contents of CH4 and NH3, indicating the gases are not derived from a high- temperature resource. Carbon-13 analyses of CO2 show a narrow range (−6.7‰ and −7.2‰) that falls within the range for MORB CO2. Thus, most CO2 seems to originate from the mantle but some CO2 could originate from thermal degradation of organic remains in underlying Cretaceous rocks. 3/4He ratios range from about 2 to 3 Rc/Ra, indicating that a small mantle/magmatic He component is present in the gases. In conclusion, the mineral waters are the surface expression of a low-temperature geothermal system of limited size that originates from the combined effects of a high regional heat flow and (possibly) the remnant heat released from subjacent basaltic andesite magma bodies that constitute the root zone of the Tuxpan shield. © 2007 Elsevier B.V. All rights reserved.

Keywords: Jungapeo; geothermal assessment; Trans-Mexican Volcanic Belt; scoria cone; Tuxpan shield; geothermal waters; mineral springs; water and gas chemistry; stable and noble gas isotopes

⁎ Corresponding author. E-mail addresses: [email protected] (C. Siebe), [email protected] (F. Goff).

1. Introduction anticline is oriented in a NNW direction. Most of the lineaments observable on satellite images, aerial photo- The town of Jungapeo is located 15 km W of the graphs, and topographic maps of the Jungapeo region small city of Zitácuaro in the eastern part of Michoacán are also oriented in this direction. Further west of this State in central Mexico. This region of pleasant indentation or “gap” extends the main part of the subtropical climate is at the southern limit of the Michoacán–Guanajuato Volcanic Field (Hasenaka and Trans-Mexican Volcanic Belt (TMVB), an active Carmichael, 1985, 1987), an area of ca. 40,000 km2 that subduction-related magmatic arc that crosses Mexico contains hundreds of scoria cones including the historic from the Pacific Ocean in the W to the Gulf of Mexico in eruption (1943–1952) of Paricutin (Luhr and Simkin, the E (Fig. 1). The main river in the study area, the Río 1993). Tuxpan, flows in a N–S direction along a narrow gorge Northwest of the study area the Late Tertiary– towards its confluence with the Río Balsas, some Quaternary Los Azufres silicic complex (Dobson and 120 km S of Jungapeo. At least ten CO2-rich thermal Mahood, 1985; Pradal and Robin, 1994) hosts an springs occur along a narrow 6-km-long strip on the extensive high-enthalpy geothermal system that has lower valley slope on the W margin of the Río Tuxpan been developed in recent decades for the production of (Figs. 1 and 2). Abundant pre-Hispanic pottery sherds electricity by Mexico's national electric power compa- found by us around the El Tular spring just outside ny, the Comisión Federal de Electricidad. East of Jungapeo, indicate that these springs did already attract Zitácuaro a large group of silicic domes form the people in remote times. While two of the springs (San Zitácuaro Volcanic Complex (Capra et al., 1997). Most José Purúa and Agua Blanca) were developed for of these domes are mantled by an extensive plinian tourism as spa-hotels in the 1950s (San José Purúa used pumice fallout deposit that reaches N5 m in thickness in to be a well-known first-class resort), all the other outcrops exposed along the new freeway near the town springs still remain poorly known and are only used by of La Dieta. Capra et al. (1997) dated this deposit by the local peasants for recreational purposes. Despite their conventional radiocarbon method at 31,350+1785/ economic potential and location within a Quaternary −1460 yr. BP (corrected for ∂13C). volcano in the TMVB, a thorough discussion of the In this part of the TMVB, young volcanism has geologic and volcanic provenance of the springs, and produced basaltic to andesitic scoria cones surrounded detailed chemical and isotope analyses of the thermal by lava shields, as well as silicic domes. Mafic to fluids have never been published. The present study is intermediate lavas typically display olivine, augite, and intended to close this gap. hypersthene phenocrysts with occasional hornblende ghosts. Plagioclase is frequently absent and quartz 2. Geologic setting xenocrysts are common (Blatter and Carmichael, 1998). Pliocene volcanism in the Jungapeo area consists of Physiographically, the Jungapeo area is located at the olivine basalt lava flows which filled valleys between trenchward front of the TMVB, right beyond the eroded mountains of older Tertiary ash flow deposits. southern limit of the Mexican Altiplano, where the The capping lava flow that forms Mesa del Campo slopes rapidly descend into the depression occupied by (MC in Fig. 3) was dated by the 40Ar/39Ar method at the Balsas River. The regional basement includes 3.776±0.007 Ma (Blatter and Carmichael, 1998). Jurassic schists and Cretaceous limestones, marls, and Andesite lavas lower in the same sequence were dated shales, which crop out in valleys and narrow canyons by Petersen (1989) using the fission track method on (Petersen, 1989; McLeod, 1989). They are overlain by zircons at 4.2±0.5 Ma and by the K–Ar method on Oligocene volcanics of the Sierra Madre Occidental hornblende at 6.0±0.7 Ma. The plateau-forming lava (Pasquaré et al., 1991) which in turn are capped by flows have been dissected since 3.8 Ma by the Río plateau-forming mafic lava flows. Younger Miocene to Tuxpan, which today runs in a narrow fault-controlled Quaternary volcanic rocks of the TMVB dominate the N–S oriented canyon. This erosional activity has lead to topography, mostly as silicic domes, scoria cones, and an inversion of the topography (due to the hardness of associated lavas (Blatter and Carmichael, 1998; Blatter the capping rocks in former valleys that are now et al., 2001). distinctive high mesas). These mesas are a characteristic West of the study area, the Tzitzio anticline feature of the landscape (Figs. 3 and 4). The vertical (Mesozoic folded marine sediments) lacks a young distance between the edge of the capping olivine basalt volcanic cover and creates a “gap” or indentation in the forming Mesa del Campo (1500 m asl) and the present volcanic front of the TMVB. The axis of the Tzitzio river bed (1000 m asl) at a location at the western foot of C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 3

Fig. 1. Sketch map showing the location of the Jungapeo mineral springs (State of Michoacán) and their position within the central part of the Trans- Mexican Volcanic Belt (TMVB, see insert). Area outlined within quadrangle is shown in Fig. 2. 4 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

Fig. 2. Jungapeo area with spring sites on the east margin of the Río Tuxpan barranca. Most of the springs occur at the basal contact of the distal western lavas forming the shield crowned by the Cerro Tapia, Cerro La Cruz, Cerro Puertecito, and El Pinal scoria cones. VA=Varelas, ES=El Saúz, EG=El Guayabo, EZ=El Zapote, SJP=San José Purúa, EA=El Aguacate, AA=Agua Amarilla, AB=Agua Blanca, EP=El Puente, CA=Cañada Azul, ET=El Tular, EC=El Concreto.

El Molcajete scoria cone (Figs. 1, 3, and 4) is 500 m, and 4A). It has a characteristic conical shape with a well- implying an average down-cutting rate of 13.3 cm/ka. preserved summit crater and an altitude of 400 m above Pleistocene volcanism is largely represented by surrounding ground. Its regular slopes are cut by several scoria cones and associated lava flows. El Molcajete meters deep gullies and it erupted a 3.5-km-long lava basaltic scoria cone is located 5.5 km to the south of flow that was emplaced along the former Tuxpan Jungapeo on the W margin of the Río Tuxpan (Figs. 1, 3, riverbed. Since then, the river has cut a new 100-m- C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 5

Fig. 3. Landsat Thematic Mapper satellite image (perspective view) showing the study area from the S toward the N. Most of the springs occur along the east margin of the Río Tuxpan (RT) at the basal contact of the lavas forming the shield surrounding the Cerro Tapia (CT), Cerro La Cruz (CC), Cerro El Puertecito (CP), and Cerro El Pinal (CEP) scoria cones. Other features mentioned in the text include ZD=Ziráhuato domes, CEZ=Cerro El Zacapendo andesite lava flow, J=Jungapeo, EM=El Molcajete scoria cone, T=Tuxpan, MO=Mesa de Ocurio, MLV=Mesa La Virgen, MC=Mesa del Campo. deep channel along the W margin of the lava flow. obtained by Blatter et al. (2001) on groundmass Assuming the previously mentioned average down- separates of the basaltic andesite lavas. Because all but cutting rate of 13.3 cm/ka, an approximate age of one of the springs considered in this study are located 0.752 Ma can be estimated. Such an old age is in near the basal contact of lavas forming the Tuxpan contradiction with its youthful morphology and must shield, several lava samples were collected for analysis therefore be considered with caution. (see next section), in order to characterize this important Cerro Las Hoyas (2470 m asl), Cerro La Cruz, and aquifer. Cerro El Pinal (Figs. 1, 2, 3, and 4B, C) form a NNW- To the E the shield lavas are overlain by the silicic oriented alignment of amalgamated scoria cones that Ziráhuato twin domes (Figs. 1 and 3). These domes are erupted along a normal fault. Cerro La Cruz has been also aligned in a NNW direction and rise up to 650 m quarried in recent years exposing 60- to 100-cm-thick high above surrounding ground. The southern dome is feeder dikes that display the same NNW direction slightly higher with an altitude of 2750 m asl. A (Fig. 5). These scoria cones crown an asymmetric shield rhyodacite sample from these domes was originally of basaltic andesite lavas. To the north the lava flows dated by Demant et al. (1975) and later corrected by only reached 3 km from their source because the pre- Blatter and Carmichael (1998) at 51,000±30,000 yr. BP. existing Cerro La Campana dome impeded their The southern dome collapsed during growth to the south flowage. The shield is widest towards the W and S, as evidenced by a small landslide deposit that underlies where the lava flows surrounded the Cerro Tapia (Fig. 3) pre-Hispanic archaeological ruins near the town of San and reached as far as 6 km from the source and plunged Felipe Los Alzati. into the ancestral Río Tuxpan, where they formed a The youngest volcanic feature in the area is the temporary dam. The river has since cut down 50 m into Zacapendo andesite lava flow (Figs. 1 and 3), which the bedrock along the margin of the shield (Fig. 3). issued from a small breached cone and flowed 4 km to Assuming an average down-cutting rate of 13.3 cm/ka, the WNW on an inclined plane until it ponded against an age of 0.376 Ma was estimated for the shield. the older Mesa La Virgen. The cone appears to be Surprisingly, this crude estimate coincides well with the located on the SSE continuation of the NNW-oriented 40Ar/39Ar ages ranging between 0.49 and 0.60 Ma fault defined by the alignment of the cones crowning the 6 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

Fig. 4. (A) Molcajete basaltic scoria cone from the E. TR=Tuxpan River canyon, MC=Mesa del Campo. Photo taken April 1991. (B) View of the Tuxpan shield from Las Anonas (SW). LF=cliff-forming lava flows, TR=Tuxpan river, SJP=San José Purúa resort, CT=Cerro Tapia scoria cone, CC=Cerro La Cruz scoria cone. Photo taken October 29, 2005. (C) View of the Tuxpan shield from the road to Carrizal (from the SSW). In the foreground is the town of Jungapeo (J), which is built on a lava flow from the Tuxpan shield crowned by the NNW–SSE aligned scoria cones Cerro La Cruz (CC), Cerro El Puertecito (CP), and Cerro Pinal (CEP). The Ziráhuato dacite domes (ZD) and Mesa de Ocurio (MO) are also shown. Photo taken January 4, 2006. All photos by Claus Siebe.

Tuxpan shield. This lava flow, as well as older Carmichael (1998). According to 40Ar/39Ar dates underlying plagioclase-free andesite lava flows was provided by them, an older flow underneath Zacapendo the focus of a recent petrologic study by Blatter and is 0.371±0.035 Ma, while a bomb from the culminating C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 7

analysis of a spring precipitate from Agua Blanca was obtained by J. Obenholzner (Vienna, Austria). Semi- quantitative X-ray diffraction analyses of Cretaceous calcareous shale fragments and Quaternary precipitates from inactive springs were determined at Los Alamos National Laboratory (LANL) using procedures described in Chipera and Bish (2002). At the spring sites, temperatures were measured with either thermometers or thermocouple probes, pH was measured with a calibrated pH meter (UNAM samples) or with pH sensitive papers (LANL samples), and flow rates were measured with a bucket or beaker and a stopwatch (Table 4). Water samples for stable isotope analysis were collected raw into 30-ml glass bottles with Polyseal caps. The isotopes were determined at Southern Methodist University (Texas), Institute for the Study of Earth and Man using standard procedures. Samples for tritium analysis were collected raw into 500-ml glass bottles with Polyseal caps and analyses obtained from the Tritium Laboratory, University of Miami (Florida) using the electrolytic enrichment method (Ostlund and Dorsey, 1977). Water samples from each site were obtained by filtering spring water (0.45 μl) brimful into two, pre- rinsed 125-ml polypropylene bottles. A few milliliters of water were then decanted from one bottle after which 10 to 20 ml of high-purity concentrated HNO3 were added to the bottle to bring the resulting pH to ≤2. Bottles were sealed with Polyseal caps. The acidified sample Fig. 5. One of the several feeder dikes of Cerro La Cruz cone was later used for cation and trace element analyses (sampling locality 04 in Fig. 2) exposed at scoria quarry. whereas the unacidified sample was used for anion analyses. Analyses conducted at UNAM were obtained scoria cone that produced the Zacapendo lava flow is using procedures outlined in Armienta and De la Cruz- 0.236±0.139 Ma. Because of the poor soil cover of this Reyna (1995) whereas those conducted at LANL were flow and its well-preserved original surface features determined by procedures listed in Werner et al. (1997, (marked flow fronts, distinct lateral margins, pressure Table 1). ridges, etc., see also Fig. 3), we consider that this flow Gas samples were collected using large polypropyl- might be much younger (maybe only a few tens of ene funnels and Teflon tubing into 300-ml Pyrex flasks thousands of years BP) than the reported age of 0.236± having double-port Teflon valves, which were filled 0.139 Ma. with approximately 100 ml of carbonate-free 4 N NaOH (Fahlquist and Janik, 1992; Goff and Janik, 2002). 3. Methods Analyses were conducted at LANL using methods described by Werner et al. (1997) and by Goff and Janik Thin sections of selected rock samples from the (2002). Aliquots of headspace gases were transferred Tuxpan shield and other young volcanoes (Zacapendo into 10-ml quartz vials using the LANL gas extraction and Molcajete) were examined by petrographic micro- system and sealed with a hydrogen torch. The contents scope at Instituto de Geofísica, Universidad Nacional of the vials were then analyzed for noble gases at the Autónoma de México (UNAM). These samples were University of Rochester (New York) by procedures also analyzed by Inductively Coupled Plasma Emission described in Poreda and Craig (1989). Aliquots of Spectrometry (ICPES) and Instrumental Neutron Acti- NaOH solutions were removed from the gas flasks into vation Analysis (INAA) for major and trace elements at 50-ml glass bottles having Polyseal caps. Carbon-13 Activation Laboratories Ltd., Ancaster, Canada. XRF analyses of the dissolved CO2 in these aliquots were 8 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 obtained from Geochron Laboratories (Massachusetts) 9702, 9709, and 9710) were collected at the distal by standard methods. margins of the lava flows near the occurrence of the mineral springs. Samples from nearby young volcanoes 4. Mineralogical and chemical composition of Zacapendo (9706 and 9708) and Molcajete (9705) were volcanic and sedimentary rocks analyzed for comparison. The location of sample sites is given in Figs. 1 and 2. Petrographic aspects of the volcanic and sedimentary rocks from the Jungapeo area were briefly described by 4.1. Petrography McLeod (1989) and Petersen (1989). The young Zacapendo scoria cone and older underlying lavas are Modal mineralogical analyses of selected samples are the only volcanic rocks in the area that have been presented in Table 1. All the volcanic rocks studied are studied in detail with modern analytical tools. Blatter basaltic in appearance, ranging from dark to medium and Carmichael (1998) provide chemical analyses of the grey and dense to vesicular (up to 8.3 vol.%). In thin Zacapendo lava flows (SiO2 =57.5–59.5 wt.%; section, rocks from the Tuxpan shield display combina- MgO=6.0–7.3 wt.%) whose phenocryst assemblage tions of olivine and clinopyroxene phenocrysts in a includes olivine, augite, bronzite, and xenocrystic quartz groundmass of plagioclase and pyroxene microlites, in a glassy groundmass of labradorite microlites and Ti- glass, and opaques. Samples from the Santa Cruz cone magnetite grains. In addition, these scientists carried out (dike and scoria) differ notably in their texture but little phase equilibria experiments and concluded that the in their mineralogical composition from samples magma equilibrated at 1050 °C and pressures around collected from distal lava flows in the main spring 1.2 kbar at H2O-saturated conditions. This means that area. The dike sample is more crystalline and has larger equilibration conditions for these hot wet plagioclase- plagioclase crystals while the scoria is characterized by free magmas could have been very shallow (∼3 km). ovoid to round vesicles and a very glassy matrix. Lava In the present study we focused on the lavas of the flows show typical flow-alignment of crystals and Tuxpan shield because all but one (El Aguacate) of the elongated vesicles. All of these features reflect differ- mineral springs occur at the basal contact of these lavas. ences in the cooling history of the various samples. In The spatial distribution of the springs indicates that the context of the present study, the most striking feature these lavas serve as shallow aquifers and hence, a close observable in rock samples is the degree of alteration of genetic relationship between them. Samples 9703 and the olivine and clinopyroxene phenocrysts (Fig. 6). 9704 (dike and scoria) were collected near the vent at the Olivines in scoria samples from the Santa Cruz cone are Santa Cruz scoria cone. Additional four samples (9701, well preserved and contain clearly discernible cubic

Table 1 Modal mineralogical analyses (N600 points counted) of representative volcanic rocks from the Jungapeo area Sample # 9701 9702 9704 9703 9705 9706 9708 Locality Agua Blanca I Agua Blanca II C. La Cruz C. La Cruz Molcajete Zacapendo Zacapendo TS-lava TS-lava TS-dike TS-scoria Scoria Lava Lava Rock type BA BA BA BA BA A A Lat. 19°28′46.5″ 19°28′24.6″ 19°31′05.5″ 19°31′05.5″ 19°24′15.7″ 19°26′09.9″ 19°25′28.1″ Long. 100°29′42.4″ 100°29′22.5″ 100°26′29.6″ 100°26′29.6″ 100°29′39.3″ 100°27′58.2″ 100°26′55.9″ Altitude 1393 m 1473 m 2330 m 2330 m 1273 m 1701 m 1711 m Qz (%) 0 0 00003.16 Ol pc (%) 4.41 3.41 5.04 1.64 4.63 0 0 Opx pc (%) 0 0 0 0 0 2.26 3.32 Cpx pc (%) 0.65 0.33 1.95 0.18 2.57 6.13 3.66 Plag pc (%) 0 0 3.41 0000 Plag mp (%) 37.09 56.91 21.14 13.64 8.23 0 0 Pyrox. mp (%) 2.61 17.89 23.09 1.82 0 10.65 9.30 Ol mp (%) 0 0 0.33 1.09 0 0.16 0 Opaque (%) 1.96 2.93 1.63 0 0.69 0.97 0.83 Groundmass (%) 53.28 18.53 43.41 81.63 83.88 79.83 79.73 Total (%) 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Vesicles (%) 0 0.49 0 8.33 5.05 0 6.67 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 9

Fig. 6. Microphotographs (10×) of olivines in lava samples taken from the Tuxpan shield at increasing distance from the vent. (A) Unaltered olivines in scoria from the upper layers of Cerro La Cruz cone (locality 03 in Fig. 2). (B) Olivine with thin iddingsite rim in sample from dike (see also Fig. 5) feeding Cerro La Cruz cone (locality 04 in Fig. 2). Small cubic Cr-spinel crystals (see arrows) in the olivines are also observable. (C) Olivine largely transformed into iddingsite in lava flow at locality 02 (Fig. 2) at ca. 6 km from the vent. (D) Olivine completely altered into iddingsite at locality 01 near the Agua Blanca spring (Fig. 2). Note that the core of the olivine has been replaced with carbonates. inclusions of Cr-spinel (Fig. 6A). Olivines in a dike alized fluids. Samples obtained from the undersaturated feeding the cone display a thin rim of light-brown non-vadose zone at high altitudes near the vent (e.g. iddingsite (Fig. 6B). Samples from distal flows in the Santa Cruz cone) are practically unaffected. This main spring area display olivines that are partly (Fig. 6C) Table 2 to totally (Fig. 6D) replaced by iddingsite (Baker and X-ray diffraction analyses of Cretaceous shale and Quaternary spring Haggerty, 1967) and other alteration products. In the precipitates from Jungapeo area, Mexico (values in %) latter case cores of olivines tend to be hollow or filled Mineral Shale Precipitate #1 Precipitate #2 Precipitate #3 with silica and carbonates. In such cases, vesicles in the (white) (gray) (orange) lava flows are also filled by carbonates and silica while Quartz 7.1 0.1 the glassy groundmass of the rock appears relatively fresh K-feldspar 2.2 and intact (Fig. 6C and D). The above features indicate Plagioclase 12.8 that the lavas forming the distal margin of the Tuxpan Mica 3.0 shield have been affected by low-temperature ground- Calcite 32.7 11.0 Smectite 33.9 water flow after their emplacement. The least stable Clinoptilolite 8.3 minerals at near-atmospheric pressures and temperatures Amorphous silica 85.9 100 85 (olivine and clinopyroxene) have been transformed into Sepiolite? 3.0 iddingsite or even partly leached out and replaced by Goethite 15 carbonates and silica at locations saturated with miner- Total 100 100 100 100 10 .Seee l ora fVlaooyadGohra eerh13(07 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C.

Table 3 Whole rock major and trace element analyses of the Tuxpan shield, Molcajete scoria cone, Zacapendo lava flow, and Agua Blanca water precipitate Analytical Detection 9701 9709 9710 9702 9704 9703 9705 9706 9708 method limits TS-lava TS-lava TS-lava TS-lava TS-dike TS-scoria Scoria Lava Lava Precipitate Sample # Agua Blanca I S. Jose Purúa El Tular Agua Blanca II C. La Cruz C. La Cruz Molcajete Zacapendo Zacapendo Agua Blanca Lat. 19°28′46.5″ 19°29′04.9″ 19°28′14.6″ 19°28′24.6″ 19°31′05.5″ 19°31′05.5″ 19°24′15.7″ 19°26′09.9″ 19°25′28.1″ 19°28′41.8″ Long. 100°29′42.4″ 100°29′38.3″ 100°29′28.3″ 100°29′22.5″ 100°26′29.6″ 100°26′29.6″ 100°29′39.3″ 100°27′58.2″ 100°26′55.9″ 100°29′40.1″ Altitude 1393 m asl 1475 m asl 1415 m asl 1473 m asl 2330 m asl 2330 m asl 1273 m asl 1701 m asl 1711 m asl 1330 m asl (%) (%) SiO2 F-ICPES 0.01 53.12 53.87 53.92 54.10 54.49 54.58 55.04 59.34 59.55 13.95 TiO2 F-ICPES 0.01 1.28 1.33 1.36 1.35 1.27 1.29 1.04 0.87 0.85 0.14 Al2O3 F-ICPES 0.01 16.03 16.75 16.95 16.46 16.61 16.58 16.65 14.83 14.63 0.74 Fe2O3t F-ICPES 0.01 5.32 8.55 8.19 8.19 8.43 8.26 7.69 5.69 5.57 65.49 MnO F-ICPES 0.01 0.31 0.13 0.12 0.37 0.12 0.12 0.13 0.09 0.09 0.05 MgO F-ICPES 0.01 3.84 3.96 3.87 3.74 5.48 5.06 6.43 6.32 6.70 0.50 CaO F-ICPES 0.01 7.78 7.44 6.90 8.35 7.26 7.12 7.04 7.45 7.33 4.21 Na2O F-ICPES 0.01 3.71 3.93 3.93 3.77 3.84 3.80 3.65 3.70 3.65 0.43

K2O F-ICPES 0.01 1.26 1.44 1.36 1.38 1.40 1.45 1.36 1.84 1.70 0.09 – 33 P2O5 F-ICPES 0.01 0.21 0.38 0.35 0.40 0.32 0.37 0.22 0.27 0.27 4.35 LOI 1.83 1.68 1.98 2.29 −0.22 0.45 −0.09 0.45 0.51 10.29 Total 94.70 99.45 98.92 100.39 98.99 99.08 99.14 100.83 100.84 100.22

(ppm) (ppm) Be F-ICPES 1 2 2 2 2 2 2 2 2 2 – Sc INAA 0.01 16.2 16.2 16.6 15.3 16.7 16.3 17.0 12.9 12.8 – V F-ICPES 1 130 148 154 146 139 142 135 111 108 – Cr INAA 0.5 119 112 145 93.3 101 93.8 178 204 237 – Co INAA 0.1 25.8 24.3 26.9 23.2 26.9 24.3 25.2 21.0 22.7 – Ni TD-ICPES 1 64 71 82 60 84 58 145 168 190 – Cu TD-ICPES 1 24 29 55 29 25 25 28 34 39 – Zn TD-ICPES 1 89 95 99 91 87 90 78 66 66 – As INAA 1 b1 b1 b1 b1 b1 b1 b1 b1 b1 – Br INAA 0.5 b0.5 0.6 b0.5 0.7 b0.5 b0.5 1.3 0.9 0.8 – Rb INAA10182220182522182021– Sr F-ICPES 1 684 665 755 707 601 615 641 1706 1666 – Y F-ICPES 1 21 24 22 23 21 21 19 13 13 – Zr F-ICPES 1 163 197 174 219 164 191 117 133 133 –

Ag TD-ICPES 0.4 b0.4 b0.4 b0.4 b0.4 b0.4 b0.4 b0.4 b0.4 b0.4 – 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C. Cd TD-ICPES 0.5 b0.5 b0.5 b0.5 0.8 b0.5 0.5 b0.5 b0.5 b0.5 – Sb INAA 0.1 b0.1 0.1 0.1 0.1 b0.1 b0.1 b0.1 b0.1 b0.1 – Cs INAA 0.2 2.5 1.1 4.4 1.9 0.4 0.8 0.4 0.5 0.5 – Ba F-ICPES 1 344 365 362 396 360 390 338 506 498 – La INAA 0.1 20.1 20.9 20.4 21.5 17.7 21.6 13.3 29.6 29.3 – Ce INAA 1 37 42 44 38 35 40 27 53 55 – Nd INAA 1 19 21 22 20 17 19 12 30 29 – Sm INAA 0.01 3.85 4.46 4.48 3.86 3.50 3.97 2.80 4.58 4.65 – Eu INAA 0.05 1.30 1.37 1.41 1.25 1.16 1.26 0.99 1.46 1.40 – Tb INAA 0.1 0.6 0.7 0.7 0.6 0.6 0.6 0.5 0.5 0.5 – Yb INAA 0.05 1.68 1.83 1.80 1.71 1.63 1.78 1.72 1.00 1.06 – Lu INAA 0.01 0.25 0.26 0.26 0.25 0.24 0.26 0.24 0.15 0.15 – Hf INAA 0.2 3.2 3.9 3.6 3.5 3.0 3.5 2.5 3.0 3.2 – Ta INAA 0.3 0.6 0.3 0.5 0.7 0.5 0.8 0.6 0.3 0.4 – Au INAA 2 b2 b2 b2 b2 b22b2 b2 b2 – Hg INAA 1 b1 b1 b1 b1 b1 b1 b1 b1 b1 – Pb TD-ICPES 5 b5 b5 b58 b566 88– Bi TD-ICPES 5 b5 b5 b56 b5 b5 b5 b5 b5 – Th INAA 0.1 1.8 2 2 1.8 1.9 2.0 1.4 2.9 2.9 – U INAA 0.1 0.6 0.6 0.4 0.4 0.6 0.6 0.4 0.9 0.9 – – 33 11 12 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 observation is corroborated by whole-rock chemical were observed. These petrographic observations on analyses (see next section). X-ray diffraction analyses of Zacapendo thin sections basically confirm those de- precipitates collected from Tuxpan shield lavas in the scribed earlier by Blatter and Carmichael (1998). spring area (Table 2) also confirm our petrographic results. Precipitates consist mostly of amorphous silica 4.2. Geochemistry of volcanic rocks (85–100%) followed by goethite (up to 15%), calcite (up to 11%) and minor amounts of sepiolite (up to 3%). In Major and trace element analyses of samples from comparison, a shale sample from the Cretaceous marine the Tuxpan shield, Molcajete, and Zacapendo volcanoes sequence underlying the Tuxpan shield lava flows and are listed in Table 3. Several samples from the Tuxpan collected uphill from the W margin of the Tuxpan River shield were collected at different localities (proximal (along the road from Agua Blanca to Las Anonas) and distal) in order to determine compositional yielded almost equal amounts of smectite (33.9%) and variability and genetic relationships. A few additional calcite (32.7%) in addition to minor amounts of analyzed samples from the Zacapendo and Molcajete plagioclase (12.8%), clinoptilolite (8.3%), quartz scoria cones are included for comparison. (7.1%), mica (3.0%), and K-feldspar (2.2%). It seems Whole-rock silica (SiO2) contents of the Tuxpan that groundwaters incorporated HCO3-ions enhancing shield range narrowly from53.1to54.6wt.%. their ability to corrode Mg–Fe-bearing silicates such as Molcajete displays a slightly higher silica content olivine contained in the overlying Tuxpan shield lava (55%) while two samples from Zacapendo are more flows. siliceous (59.3% to 59.6%). The rocks are classified Sample 9705 from Molcajete scoria cone is similar in based on SiO2 and total alkali (Na2O+K2O) contents composition to Tuxpan shield samples. Idiomorphic (Fig. 7). All of the analyzed rocks are subalkaline. millimeter-sized olivine crystals with Cr-spinel inclusions According to this scheme, the Tuxpan shield and occur together with augite in a glassy matrix with feldspar Molcajete cone produced calc–alkaline basaltic ande- microlites. In strong contrast, Zacapendo samples are sites while Zacapendo erupted high-silica andesites. characterized by orthopyroxene and clinopyroxene phe- Comparison of SiO2 and MgO contents in distal nocrysts set in a glassy matrix that includes pyroxene and lavas and proximal samples (dike and scoria) of the feldspar microlites. Clinopyroxene phenocrysts either Tuxpan shield indicate a clear trend. Distal lavas have occur as single individuals or in glomeroporphyritic the lowest SiO2 contents while proximal samples are clusters. In addition, hornblende “ghosts” (totally altered more silicic. Surprisingly, MgO contents are highest in to oxides) as well as millimeter-sized quartz xenocrysts the most silicic proximal samples (MgO=5.0 to 5.5%)

Fig. 7. Total alkalis (Na2O+K2O) plotted against silica (SiO2) after Les Bas et al. (1986) for all analyzed Jungapeo area volcanic rocks. Analyses of samples from Pelado, Guespalapa, and Chichinautzin scoria cones in the Sierra Chichinautzin Volcanic Field near Mexico City (from Siebe et al., 2004) and additional samples from Zacapendo (from Blatter and Carmichael, 1998) are shown for reference. The dividing line separating alkali olivine basalts and tholeiites in Hawaii (MacDonald and Katsura, 1964) is also shown. C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 13

Fig. 8. Variation diagrams showing selected major (in wt.%) and minor elements (in ppm). Key symbols as in Fig. 7. and lowest in the least silicic distal samples (MgO=3.7– Different major and trace elements were plotted 4.0%) (see Table 3 and Fig. 8A). This apparent against silica as well as other elements (Figs. 7 and 8). contradiction is easily explained by the strong iddingsi- Analyses of basaltic to andesitic samples from mono- tization of the olivines observed in thin sections of distal genetic Guespalapa, Chichinautzin, and Pelado scoria samples. In these samples Mg has been leached out by cones located within the Sierra Chichinautzin Volcanic low-temperature thermal waters replacing the olivines Field south of Mexico City (Siebe et al., 2004) were by Fe-oxides, silica, and carbonates (Fig. 6). Distal lavas included for comparison. From these plots it becomes also display high (1.6–2.3 wt.%) loss on ignition (LOI) clear that rocks from the Tuxpan shield and Molcajete values indicating the presence of hydrated glass and bear many resemblances with rocks from the Sierra other secondary minerals (Table 3). Chichinautzin. On the other hand, Zacapendo high-Si 14 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

conclusion before (Siebe et al., 2004; Schaaf et al., 2005). The close spatial occurrence of monogenetic volcanoes, whose origin can be chemically traced back to different deep sources in the mantle, implies rapid ascent along extensional faults of relatively small batches of magma. Rapid ascent without stagnation in larger shallow magma chambers prevents the induction of high-temperature geothermal systems. From the above arguments, the occurrence of larger subvolcanic bodies of cooling magma seems unlikely underneath the Jungapeo area.

5. Geohydrology and visual characteristics of springs

The Tuxpan shield from which the majority of springs issue forms a broad cone having a radius from the summit to the west of more than 6 km. The area occupied by the Tuxpan shield is 92 km2 (determined using a compensating polar planimeter on topographic maps). An accurate thickness of the shield is difficult to determine because the pre-existing topography is not known. None the less, the minimum volume of erupted products was crudely estimated to be 6.44 km3 by 2 Fig. 9. (A) Chondrite-normalized (Sun and McDonough, 1989) REE multiplying the area covered by the shield (92 km ) with compositions (in ppm) of analyzed volcanic rocks from the Jungapeo an average thickness (70 m) determined by observations area. (B) Trace element abundances normalized to primitive mantle of at lava flow margins along the Tuxpan River canyon. analyzed volcanic rocks from the Jungapeo area. Primitive mantle The nearest meteorological station to Jungapeo is at values employed are from Sun and McDonough (1989). Laguna del Fresno near Maravatío, about 55 km to the north. From 1992 to 2005, the average yearly rainfall for andesites are quite unusual with extraordinarily high 12 years at this station is 700±127 mm/yr (±18%) (see MgO, Sr, and Ni contents (Fig. 8A, B, and D, also Fig. 10 for more details). Using the computed value respectively). Although much more evolved than for average rainfall and a catchment area for rainfall on Tuxpan shield lavas (see also the Rare Earth element the shield of 92 km2, the annual volume of water patterns and trace element distributions in Fig. 9), recharged into the shield is about 6.5±1.3×108 l/yr. Zacapendo rocks are higher in certain compatible This recharge water percolates into the permeable elements such as Mg and Ni. This indicates that Tuxpan basaltic andesite lava flows and flow breccias of the shield compositions were not parental to Zacapendo shield and flows radially away from the summit down lavas and therefore excludes a simple genetic relation- hydrologic gradient. Much of this water emerges as ship by crystal fractionation processes. This conclusion springs around the western flank of the shield and most is quite striking, especially if one considers that both, the of the springs consist of the CO2-rich thermal waters Tuxpan shield and Zacapendo lava flow are closely along the Río Tuxpan in the Jungapeo area. The related in time and space and maybe even located on the estimated yearly discharge volume of these thermal same fault at a horizontal distance of less than 8 km springs is roughly 6×108 l/yr based on flow measure- (Fig. 1). Without getting into too many petrogenetic ments obtained in March 1997 (Table 4). The discharge details (which would be beyond the scope this study), it measurements were measured with a bucket and can be said that chemical analyses of Tuxpan and stopwatch in most cases having an error of ±10%, but Zacapendo lavas point toward a heterogeneous do not account for seasonal variations. Yet it seems (enriched/depleted±hydrated) lithospheric mantle apparent that much of the water recharged on the shield source underneath this part of the TMVB. The detailed can be accounted for in the discharge from the springs. study of monogenetic volcanism in other parts of the Many of the springs are unimproved issuing from TMVB (e.g. Sierra Chichinautzin) has lead to such a natural or crudely constructed pools in basaltic andesite C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 15

Fig. 10. Average monthly precipitation and average monthly minimum, mean, and maximum temperatures for the period 1992–2005 as measured at Laguna del Fresno, located 7 km south of Maravatío and 55 km N of Jungapeo. rocks along the eastern canyon wall (Fig. 11A). Fe-oxides as the waters cascade toward the Río Tuxpan. However, the springs at SJP, AB, EA, and CA are An old leaflet distributed to tourists several decades ago improved with emergence points protected by grottos for the purpose of advertising the medical benefits of the and concrete boxes, and flow directed through pipes into mineral waters, included a paragraph claiming that the various types of commercialized and private pools elevated levels of radioactivity of the waters would also (Fig. 11B to D). The dominating characteristics of the contribute to their soothing effects. As a result, we thermal waters are their moderate temperatures (28 to measured radioactivity with a Geiger–Müller detector 32 °C), relatively high flow rates, effervescence of CO2 but found that the spring waters and surrounding rocks gas, clarity at emergence, and precipitation of hydrous are not particularly radioactive and are indistinguishable iron oxides around pool margins and channel ways. from the regional background noise. Waters contained in larger pools become pale green to An analysis of the dried Fe-rich coagulant from the orange as dissolved ferrous iron converts to coagulants main pool at Agua Blanca (Table 3) reveals about 65% of hydrated ferric oxides (Fig. 11A, B, and C). Addi- Fe2O3, 14% SiO2, 4% CaO, and 4% P2O5 (by weight). tional precipitates of carbonate and silica form with the X-ray diffraction analysis of this sample shows that it is 16

Table 4 a Field and isotope data for CO2-rich spring waters, Jungapeo area, Mexico

Sample Location Latitude Longitude Altitude Date Temperature Field pH Flow δ18 O δD 3H Rock type 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C. (m) (°C) (l/min) (‰) (‰) (TU) MVT-3 San José Purúa b 19°29′19.8″ 100°29′31.6″ 1455 24/06/1995 31.8 6.2 nd −8.7 −67 0.82±0.09 Basaltic andesite MVT-27 San José Purúab 19°29′19.8″ 100°29′31.6″ 1455 03/03/1997 32.4 6.0 150 −9.2 −61 nd Basaltic andesite MVT-4 Agua Blanca; resort c 19°28′41.8″ 100°29′40.1″ 25/06/1995 31.0 6.3 nd −8.7 −66 1.33±0.09 Basaltic andesite MVT-25 Agua Blanca; source d 19°28′47.4″ 100°29′34.2″ 1340 02/03/1997 29.7 6.0 N200 −9.2 −65 1.22±0.09 Basaltic andesite EP-9403 El Puente e 19°28′45.9″ 100°29′38.1″ 1335 01/03/1994 29.1 6.0 nd nd nd nd Basaltic andesite MVT-22 Agua Amarilla f 19°29′06.7″ 100°29′48.3″ 1350 01/03/1997 30.6 6.2 300 −9.3 −63 nd Basaltic andesite MVT-23 El Tular g 19°28′09.0″ 100°29′32.1″ 1315 02/03/1997 29.2 6.0 60 −9.4 −63 0.33±0.09 Basaltic andesite MVT-24 Cañada Azul; pool h 19°28′32.3″ 100°29′32.3″ 1355 02/03/1997 26.0 6.0 ≥10 −9.3 −63 1.21±0.09 Basaltic andesite MVT-26 El Aguacate; pool i 19°29′09.7″ 100°29′50.4″ 1300 02/03/1997 31.8 6.5 15 −8.8 −64 1.78±0.09 Limestone MVT-28 Varelas j 19°30′22″ 100°29′39″ 1475 03/03/1997 29.8 5.8 250 −9.6 −68 nd Basaltic andesite MVT-29 El Sauz k 19°30′06″ 100°29′34″ 1500 03/03/1997 28.3 5.7 80 −9.6 −62 nd Basaltic andesite MVT-30 El Gauyabo l 19°30′06″ 100°29′28″ 1510 03/03/1997 29.4 5.5 60 −9.5 −61 nd Basaltic andesite MVT-31 El Concreto m 03/03/1997 24.5 6.8 120 −9.5 −64 2.09±0.11 Basaltic andesite MVT-32 El Zapote n 19°30′03″ 100°29′24″ 1525 03/03/1997 30.2 6.0 15 −9.3 −64 0.69±0.09 Basaltic andesite a Temperature by probe; pH by indicator strips; flow by visual estimate; oxygen-18 error=±0.25‰; deuterium error=±1‰; tritium error is shown in table. b From pool in grotto constructed beneath cliff of lava; free gas; Fe-oxide precipitate; other springs on property, some possibly warmer. c From inlet pipe to main pool; considerable gas discharge and Fe-oxide precipitate. d From beneath concrete box around source in ravine of lava and flow breccia; large amount of gas; Fe-oxide precipitate. e From abandoned pool just upstream of bridge on road; minor gas; minor Fe-oxide precipitate. f From source at rock pool in colluvium about 6 m below contact with lava; free gas; considerable Fe-oxide precipitate. g From source above rock pool in lava talus; free gas; considerable Fe-oxide precipitate. h From source in concrete box in lava; free gas; considerable Fe-oxide precipitate. i From terrace gravel and colluvium on southwest side of Rio Tuxpan; diluted with river water or groundwater; minor gas; faint H2S odor. j From source at rock pool in lava; very minor gas; considerable Fe-oxide precipitate. – 33 k From rock pool in colluvium beneath lava; no gas; minor Fe-oxide precipitate. l From rock pool in colluvium beneath lava; free gas; minor Fe-oxide precipitate. m From concrete box in colluvium; collected as cold background sample. n From rock pool in colluvium beneath contact with lava; free gas; minor Fe-oxide precipitate. C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 17

Fig. 11. (A) Small bathing pool at El Tular mineral spring near Jungapeo. Note characteristic yellowish color due to Fe-mineral precipitation. Photo taken August 1992. (B) Agua Blanca resort with swimming pools. Mineral water is conducted in a tube from a spring located 100 m behind the viewer to the smaller (greenish) pool and from there to the larger (yellow) pool. Differences in color are due to increasing amounts of precipitates at water temperatures ranging from 31 °C (green pool) to 28 °C (yellow pool). Photo taken February 28, 2004. (C) San José Purúa resort. Note characteristic yellow color due to Fe-mineral precipitation. Photo taken October 28, 2005. (D) El Aguacate mineral spring located on the west margin of the Río Tuxpan. During the rainy season (June–September) this spring is often inundated by the torrent of the river. Photo taken April 1994. All photos by Claus Siebe. 18

Table 5 Selected major and trace element analyses of CO2-rich waters from Jungapeo area, Mexico (values in ppm unless otherwise noted; na means not analyzed or measured). A complete tabulation of our 85 major element analyses of the Jungapeo waters can be obtained from the first author

Sample Location Date (mm/dd/yyyy) Analyst Temperature (°C) pH (lab) SiO2 Na K Li Ca Mg Fe F Cl HCO3 SO4 B TDS SJP-9408 San José Purúa 11/08/1994 Armienta 31.7 6.77 114 235 21.4 na 114 84.9 7.34 na 141 1310 10.7 7.07 1994 .Seee l ora fVlaooyadGohra eerh13(07 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C. SJP-9410 San José Purúa 09/10/1994 Armienta 32.3 6.73 117 279 31.4 na 111 80.1 7.17 0.56 145 1285 10.4 6.86 2073 MVT-3 San José Purúa 24/06/1995 Counce 31.8 6.85 143 257 23.1 1.39 120 92.2 7.33 0.21 148 1195 10.2 7.53 2008 MVT-27 San José Purúa 03/03/1997 Counce 32.4 7.82 136 243 22.3 1.36 109 81.8 6.11 0.19 151 1205 9.63 8.10 1975 AB-9408 Agua Blanca 11/08/1994 Armienta 29.5 6.34 107 225 17.7 na 115 75.7 7.83 na 139 1240 12.9 7.04 1876 AB-9410 Agua Blanca 08/10/1994 Armienta 30.0 6.77 96.1 245 24.8 na 111 79.2 7.86 0.347 130 1065 13.3 5.66 1778 MVT-4 Agua Blanca 25/06/1995 Counce 31.0 6.82 130 218 19.1 1.10 110 84.5 7.23 0.30 134 1095 12.7 7.00 1822 MVT-25 Agua Blanca 02/03/1997 Counce 29.7 6.99 127 219 20.5 1.21 107 78.6 6.87 0.23 139 1110 13.5 7.61 1832 EP-9406 El Puente 18/06/1994 Armienta 28.9 6.19 102 175 19 na 109 66.0 na na 105 1030 20.3 8.26 1635 EP-9410 El Puente 08/10/1994 Armienta 29.1 6.57 70.2 194 17.7 na 96.2 73.4 7.38 0.45 105 1060 20.2 5.80 1650 AA-9408 Agua Amarilla 18/06/1994 Armienta 30.8 6.43 165 217 20.8 na 109 78.2 5.4 na 140 1175 20.0 6.01 1877 AA-9410 Agua Amarilla 08/10/1994 Armienta 30.8 6.8 86.9 228 17.3 na 105 77.4 4.36 0.32 125 1140 19.9 5.97 1810 MVT-22 Agua Amarilla 01/03/1997 Counce 30.6 7.21 129 212 19.6 1.15 105 81.7 4.86 0.27 128 1060 21 6.73 1772 ET-9408 El Tular 13/08/1994 Armienta 28.4 6.37 104 182 17.6 na 96.7 67.8 8.0 na 107 937 3.15 5.59 1529 ET-9410 El Tular 08/10/1994 Armienta 29.7 6.51 72.5 196 23.2 na 102 64.9 6.89 0.35 105 1005 2.76 5.94 1585 MVT-23 El Tular 02/03/1997 Counce 29.2 6.85 123 176 17.3 0.93 90.2 64.8 5.45 0.27 110 929 3.75 6.36 1530 CA-9408 Cañada Azul 11/08/1994 Armienta 27.0 6.09 93.6 146 14.5 na 79.6 58.0 4.74 na 84.5 770 12 4.63 1268 CA-9410 Cañada Azul 08/10/1994 Armienta 27.3 6.21 91.6 161 20.8 na 78.6 56.3 4.94 0.4 85 831 11.9 4.32 1346 MVT-24 Cañada Azul 02/03/1997 Counce 26.0 6.89 115 144 15.1 0.75 79.6 60.7 4.16 0.25 90.8 792 11.7 4.76 1323 EA-9404 El Aguacate 01/04/1994 Armienta 34.7 6.69 96.5 1425 60 na 72.8 50.4 0.86 0.65 660 2290 98.3 29.1 4784 EA-9406 El Aguacate 18/06/1994 Armienta 33.8 6.48 70.6 750 35 na 77.3 47.1 na na 500 1570 80.7 24.1 3155 MVT-26 El Aguacate 02/03/1997 Counce 31.8 7.92 86.5 826 30.6 2.03 88.5 32.7 0.10 0.44 446 1877 96.2 19.7 3512 VA-9408 Varelas 12/08/1994 Armienta 29.6 6.45 103 118 11.0 na 109 23.2 5.91 na 42.6 755 16.2 1.74 1148 VA-9410 Varelas 09/10/1994 Armienta 30.1 6.38 95.1 137 13.5 na 66.8 48.8 5.2 0.35 43.0 731 16.4 3.84 1161 MVT-28 Varelas 03/03/1997 Counce 29.8 6.96 126 112 11.5 0.47 63.0 45.1 4.86 0.30 42.8 674 15.2 2.29 1099 ES-9408 El Sauz 12/08/1994 Armienta 28.1 6.48 99.3 88.6 8.74 na 48.3 40.9 4.39 na 27.1 523 19.9 1.33 862 ES-9410 El Sauz 09/10/1994 Armienta 28.8 6.45 103 99.6 12.9 na 47.1 40.5 3.93 0.45 30.0 545 20.4 1.67 905 – MVT-29 El Sauz 03/03/1997 Counce 28.3 7.51 118 83.4 9.31 0.35 46.2 38.4 3.90 0.34 28.1 509 20.1 1.54 860 33 EG-9410 El Gauyabo 09/10/1994 Armienta 30.0 6.11 74.0 206 13.9 na 73.2 61.9 7.26 0.32 49.5 827 7.67 2.44 1323 MVT-30 El Gauyabo 03/03/1997 Counce 29.4 7.44 126 115 12.5 0.51 70.2 60.1 5.90 0.28 48.3 765 8.49 2.66 1217 MVT-31 El Concreto 03/03/1997 Counce 24.5 8.19 105 60.6 8.66 0.23 39.5 47.7 0.04 0.33 29.5 445 41.7 1.30 757 EZ-9408 El Zapote 12/08/1994 Armienta 29.1 6.47 106 136 15.1 na 84.6 71.9 6.23 na 63.0 965 10.3 3.03 1412 EZ-9410 El Zapote 08/10/1994 Armienta 29.8 6.50 93.2 148 16.1 na 77.6 76.2 5.22 0.37 62.5 918 10.8 2.93 1411 MVT-32 El Zapote 03/03/1997 Counce 30.2 7.77 131 131 14.3 0.59 80.3 72.3 5.29 0.24 61.8 864 10.7 3.21 1378 .Seee l ora fVlaooyadGohra eerh13(07 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C.

Table 6 Trace element analyses of thermal waters from Jungapeo area, Mexico (values in ppm unless otherwise noted). Footnotes same as Table 4 a

Sample Location Analyst Temperature (°C) Al As Ba Br Cs Cu I Mn NH4 Ni NO3 PO4 Rb Sr V SJP-9408 San José Purúa Armienta 31.7 0.1 b0.001 na 0.3 na na na na na na 2.4 b0.05 na 0.75 na MVT-3 San José Purúa Counce 31.8 0.12 0.002 0.29 0.31 0.21 b0.002 b0.01 0.36 1.02 0.009 0.14 b0.05 0.15 1.02 0.002 MVT-27 San José Purúa Counce 32.4 0.38 0.0020 0.30 0.25 0.14 0.002 0.010 0.33 1.02 b0.002 0.04 b0.05 0.13 0.98 0.002 AB-9408 Agua Blanca Armienta 29.5 0.05 0.001 na 0.27 na na na na na na b0.02 na na 0.5 na MVT-4 Agua Blanca Counce 31.0 0.17 0.001 0.28 0.27 0.20 b0.002 b0.05 0.31 1.16 0.003 0.04 b0.05 0.11 0.93 b0.002 MVT-25 Agua Blanca Counce 29.7 0.40 0.0032 0.30 0.20 0.18 b0.002 0.008 0.29 0.66 0.007 b0.02 b0.05 0.13 0.97 0.002 AA-9408 Agua Amarilla Armienta 30.8 0.06 b0.001 na 0.26 na na na na na na 0.2 b0.05 na 0.6 na MVT-22 Agua Amarilla Counce 30.6 0.41 0.0013 0.27 0.18 0.18 0.003 b0.005 0.37 0.89 0.006 0.03 b0.05 0.15 0.93 b0.002 MVT-23 El Tular Counce 29.2 0.33 0.0006 0.23 0.19 0.13 0.002 0.005 0.28 0.76 b0.002 b0.02 b0.05 0.11 0.77 b0.002 MVT-24 Cañada Azul Counce 26.0 0.23 0.0007 0.18 0.13 0.15 b0.002 0.007 0.24 0.57 b0.002 2.15 b0.05 0.11 0.65 0.004 MVT-26 El Aguacate Counce 31.8 0.33 0.037 0.11 0.70 0.13 0.009 0.017 0.18 0.62 b0.002 2.29 b0.05 0.12 1.76 0.004 VA-9408 Varelas Armienta 29.6 0.1 b0.001 na 0.05 na na na na na na 0.78 b0.05 na 0.39 na MVT-28 Varelas Counce 29.8 0.34 0.0014 0.12 0.11 0.092 b0.002 0.017 0.24 0.46 b0.002 0.02 b0.05 0.069 0.52 0.003 MVT-29 El Sauz Counce 28.3 0.20 0.0036 0.07 0.05 0.062 0.002 0.009 0.18 0.29 0.006 0.04 b0.05 0.049 0.36 0.002 MVT-30 El Gauyabo Counce 29.4 0.22 0.0019 0.12 0.09 0.089 0.036 0.006 0.32 0.38 0.054 0.11 b0.05 0.074 0.59 0.002 MVT-31 El Concreto Counce 24.5 0.13 0.0010 0.02 0.04 0.015 0.003 b0.005 0.002 0.06 0.006 7.10 0.6 0.036 0.34 0.033 EZ-9408 El Zapote Armienta 29.1 0.1 b0.001 na b0.02 na na na na na na 2.2 b0.05 na 0.48 na MVT-32 El Zapote Counce 30.2 0.28 0.0030 0.12 0.15 0.093 b0.002 b0.005 0.35 0.32 0.010 1.09 b0.05 0.090 0.66 0.013 a All samples were at or below detection limits for the following elements: Ag≤0.002; Beb0.002; Cdb0.001; Cob0.002; Crb0.002; Hgb0.0002; Mo≤0.002; NO2 ≤0.02; Pb≤0.002; Sbb0.0002; Se≤0.0002; S2O3 b0.01; Ti≤0.01; Zn≤0.02. – 33 19 20 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

solution is Fe2+). Thus, Jungapeo mineral waters do not have the typical chemical composition of high-temperature geothermal reservoir fluids or derivative hot springs (Goff and Janik, 2000). Instead, the Jungapeo fluids resemble soda spring waters from other parts of the world that may or may not be associated with high- to moderate- temperature geothermal activity (e.g. Barnes et al., 1973; Pauwels et al., 1997; Marques et al., 2004; Inguaggiato et al., 2005; Choi et al., 2005). Waters from El Aguacate (EA) are somewhat different from those mentioned above. Waters from EA have much more Na+K, Cl, B, and As but much less Ca+Mg and Fe+Mn than other Jungapeo fluids. As a result EA waters appear more geothermal in character than other Jungapeo fluids. On a bicarbonate–chloride–sulfate ternary (Fig. 12), Jungapeo waters plot in the immature field of geothermal Fig. 12. Triangular plot of HCO /10-SO -Cl for waters of the Jungapeo 3 4 fluids as defined by Giggenbach (1992).NotethatHCO3/ area, Mexico. The plot is modified from Giggenbach (1992) to provide 10 is used on this plot so that the Jungapeo data do not more definition for bicarbonate-rich waters. form a clump of points near the bicarbonate apex. Thus, El Aguacate (EA) fluids are the least immature whereas El amorphous. Assuming that CaO is actually CaCO3, Concreto (EC), El Gauyabo (EG), El Sauz (ES), El Zapote carbonate probably makes up 7% of the sample. Ferric (EZ), and Varelas (VS) waters are the most immature. oxides, calcite, and opaline silica can be found as Two trends are discernible on the plot, both with EZ and cements and coatings in the basaltic rocks near the EG waters at one end: a trend of relatively increasing spring areas, especially in the flow breccias. On the western margin of the shield south of San José Purúa abundant fragments of opaline silica and calcite with coatings of Fe-rich oxides are found in the cultivated surface of the basalt suggesting that thermal springs may have once discharged at higher elevations. Road cuts through soils into the underlying basalt commonly reveal deposits of caliche as much as 0.5 m thick formed at the base of the soil horizon. X-ray diffraction analyses of three samples of these old deposits (Table 2) verify that they consist primarily of opal-A to opal-CT, goethite and calcite. One sample possibly contains a small amount of sepiolite (Mg4Si6O15(OH)2(H2O)6).

6. Water geochemistry

6.1. Chemistry and mixing relations

Selected major element analyses of Jungapeo waters are presented in Table 5 while all trace element data of waters are listed in Table 6. From a geothermal perspective, Jungapeo mineral waters are chemically characterized by moderate SiO2, Ca+Mg nearly equal to Fig. 13. Plot of (Fe+Mn) versus HCO3 for waters of the Jungapeo Na+K, high HCO3, moderate to low Cl, and low F and area, Mexico. Most Jungapeo CO2-rich waters form a single trend of SO4. With respect to key geothermal trace elements, B is high (Fe+Mn) with increasing bicarbonate. Water from El Aguacate rather high, Li is moderate, while Br and As are quite low. spring is an exception containing much less (Fe+Mn) relative to In contrast, Fe+Mn is exceptionally high (the iron in bicarbonate. C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 21

Fig. 14. Chloride variation diagrams using As, B, Br, Li, Cs, and Rb for waters of the Jungapeo area, Mexico. The Cl:Br line for seawater is calculated from Krauskopf (1979). 22 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

(relatively low Fe+Mn) discharge from Cretaceous rocks W of the Río Tuxpan. Because they are glassy, the basaltic andesite rocks are highly reactive to bicarbonate-rich fluids and probably release their Fe+Mn rather easily, whereas the Cretaceous marls contain less Fe+Mn and release less of these constituents into solution. In this plot, the sample from EC (the sample with the least Cl) contains virtually no Fe+Mn suggesting that EC is composed of very young groundwater that interacts little with host rocks. A plot of As versus Cl (Fig. 14) shows that most Jungapeo fluids plot as a cluster of very low As with respect to Cl. As does not increase as Cl increases. The exception is water from EA, which contains substantially higher As than other Jungapeo waters, although not nearly as high as As in most high-temperature geothermal fluids (see Table 3 of Goff and Janik, 2000). The B versus Cl plot (Fig. 14) shows a very strong trend of increasing B with Cl. There are two segments

Fig. 15. Plot of Mg and Cl vs. sampling date for waters discharged from two springs in the Jungapeo area, Mexico. chloride and decreasing bicarbonate toward mature waters and a trend of nearly constant Cl/HCO3 but increasing sulfate toward the composition of EC water. The first trend suggests possible mixing between bicarbonate-rich groundwaters and a deep geothermal end-member, while the second trend indicates mixing of bicarbonate-rich groundwaters with a source of sulfate. In this case, EC water is also enhanced in nitrate and phosphate (Table 6) suggesting that the excess sulfate comes from agriculture (fertilizers such as ammonium sulfate, etc.) and not from a Fig. 16. Plot of ∂D versus ∂18O for waters of the Jungapeo area, deep geothermal or volcanic source. Mexico (enclosed by solid line). WMWL=World Meteoric Water Line, ∂D=8∂18O+10 (Craig, 1961), MBWL=Mexico City Basin On the plot of Fig. 13, most Jungapeo fluids form a meteoric water line (dashed), ∂D=7.97∂18O+11.03 (Cortés et al., positive trend of highly increasing Fe+Mn with HCO3.In 1997), LAMW=oval showing range of Los Azufres meteoric waters contrast, EA waters form a completely separate trend of and LAGW=range of Los Azufres pre-production geothermal well waters (Torres-Alvarado et al., 1995; Barragán et al., 2005). Additional slightly increasing Fe+Mn with increasing HCO3.The difference between the two trends is explained by ovals are also shown for the majority of Mexico City Basin meteoric waters as defined by the Gaussian distribution of data in Cortés et al. differences in the host rocks. Waters with high Fe+Mn (1997, p. 374) and for the Popocatépetl data set of Inguaggiato et al. contents discharge from young glassy basaltic andesite (2005). Plus symbols show data from Popocatépetl reported in Werner rocks of the Tuxpan shield volcano whereas the EAwaters et al. (1997), and F. Goff (unpublished data). C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 23 on the trend: a concentrated segment (from a deeper Mg) from the two types of mineral water: less source) for EA waters and a less concentrated segment concentrated fluid represented by Agua Blanca spring (from a shallower source) for other Jungapeo waters. and the more concentrated water of El Aguacate spring. Within both segments, the data indicate that mixing with Chloride is the more conservative (soluble) component dilute water affects the chemistry of the two types of while Mg is a less conservative component, partially fluids. The Br–Cl plot (Fig. 14) mimics the B–Cl plot, controlled by carbonate solubility. We obtained 14 water defining one trend with a concentrated fluid represented samples from AB during the 6-year period from early by EA and a trend with less concentrated fluids 1991 to early 1997. Allowing that some of the variations represented by other Jungapeo fluids. The Jungapeo are due to analytical errors, we observe that there are Br–Cl trend is not the same as the trend for seawater, fluctuations in chemistry that probably represent sea- which would imply that marine rocks are not the sole sonal dilution of mineral water with dilute groundwater source of Br. Thus, Cl and Br may come from a variety and/or rainwater. Thus, in very early 1994, dry of sources. conditions in the area caused the Cl content of AB Plots of Li, Cs, and Rb vs. Cl (Fig. 14) all show the water to rise to levels of roughly 170 ppm but wet same pattern: a well-defined trend for less concentrated conditions commencing in April caused a reduction of Cl mineral fluids issuing from the basaltic andesite shield in AB water to more typical Cl values of roughly volcano and a different cation–chloride relationship for 130 ppm. Note that Mg content apparently behaves concentrated fluid represented by EA. EA waters flow independent of Cl concentration and this independent through Cretaceous rocks on the W side of the Río behavior is evident in all springs. Only five samples Tuxpan. Possibly, Li, Cs, and Rb are more easily comprise the EA data set (Fig. 15) but Cl content is dissolved in carbonated waters circulating in glassy positively correlated with temperature. Again, climate basaltic andesite rocks than in marine Cretaceous rocks. variations related to wet and dry seasons apparently In contrast, rock type does not apparently affect the affect the concentration of Cl in EA. Mg is less affected relative solubility of B and Br. by climate than Cl. Fig. 15 compares concentration vs. sampling date for two constituents of different chemical behavior (Cl and 6.2. Recharge to system

The plot of deuterium vs. oxygen-18 (Fig. 16) shows that the CO2-rich thermal waters of the Jungapeo region form a cluster of points straddling the World Meteoric Water Line (WMWL, Craig, 1961) and the Mexico City Basin meteoric water line (MBWL, Cortés et al., 1997). Thus, Jungapeo thermal waters are recharged by precipitation of local meteoric water as indicated by our recharge–discharge measurements in Section 5. Cold water from El Concreto shows that background water and thermal waters are essentially identical in isotopic composition indicating that the Tuxpan shield is the recharge area for most springs. El Aguacate, the only spring discharging on the W side of the Río Tuxpan, has similar isotope composition to the rest of the springs. Thus, El Aguacate is also composed mostly of meteoric water from a similar local source even though water from this spring is substantially more concentrated in chloride than the others (see above). For comparisons, the dashed ovals outline the areas of cold meteoric waters in the Los Azufres geothermal area (Torres-Alvarado et al., 1995), only 50 km NW of Jungapeo, for the majority of cold meteoric waters in the Fig. 17. Plot of 3H versus Cl for waters of the Jungapeo area, Mexico. Mexico City Basin (Cortés et al., 1997), and for a small Rain data from 1985 to 1992 are from Goff et al. (1987), Janik et al. group of cold meteoric waters collected from the region (1992) and Goff and McMurtry (2000). around Popocatépetl Volcano (Inguaggiato et al., 2005) 24 .Seee l ora fVlaooyadGohra eerh13(07 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C.

Table 7 Gas analyses from selected bicarbonate-rich waters, Jungapeo, Mexico (values in mol%, unless noted) a 13 Sample # Description Date (mm/ Temperature H2 Ar O2 N2 CH4 C2H6 CO CO2 NH3 H2S Total δ C–CO2 dd/yyyy) (°C) (‰ PDB) MVT-3 San José Purúa Spring, 24/06/1995 31.8 b0.0003 0.0159 0.3325 1.0476 0.00376 b0.0003 b0.0002 98.61 0.00128 b0.001 100.01 −6.7⁎ gazebo with Virgin MVT-27 San José Purúa Spring, 03/03/1997 32.4 b0.001 0.0781 2.0412 7.2992 b0.002 b0.002 b0.001 90.59 b0.001 b0.00004 100.01 gazebo with Virgin MVT-4 Agua Blanca Spring, pipe 25/06/1995 31.0 b0.00004 0.0045 0.0195 0.1875 0.00865 b0.00004 b0.00002 99.78 0.00086 b0.001 100.00 −7.2 in concrete box @ source MVT-25 Agua Blanca Spring, pipe 02/03/1997 29.7 b0.00005 0.0037 0.0132 0.174 0.0085 b0.0001 b0.00005 99.8 b0.0005 b0.00001 100.00 −7.0 in concrete box @ source MVT-22b Agua Amarilla Spring, 01/03/1997 30.6 b0.002 0.1127 2.8331 11.5492 b0.003 b0.003 b0.002 85.55 b0.002 b0.00005 100.05 from concrete box MVT-23 El Tular Spring 02/03/1997 29.2 b0.0002 0.0274 0.1039 1.2715 0.01252 b0.0003 b0.0002 98.59 b0.0005 b0.00002 100.01 −6.8 MVT-24 Cañada Azul Spring, 02/03/1997 26.0 b0.0002 0.0167 0.276 1.0592 0.00666 b0.0003 b0.0002 98.64 b0.0006 b0.00002 100.00 −7.0⁎ from concrete box MVT-26b El Aguacate Spring 02/03/1997 31.8 b0.0005 0.0504 1.2426 3.9409 b0.001 b0.001 b0.0005 94.8 b0.001 b0.00003 100.03 −6.8 MVT-30 El Guayabo Spring 03/03/1997 29.4 b0.002 0.1441 3.0069 12.3995 b0.003 b0.003 b0.002 84.5 b0.002 b0.00007 100.05 MVT-32 El Zapote Spring 03/03/1997 30.2 b0.0002 0.027 0.2077 1.3515 0.01805 b0.0003 b0.0002 98.4 b0.0006 b0.00002 100.00 −7.0⁎ a Detection limits vary depending on the date of analysis and standardization. Carbon isotope values with asterisk are average of two analyses. – 33 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 25

Table 8 Noble gas analyses and gas ratios on samples from the Jungapeo area, Mexico 3 4 3 4 4 40 36 20 36 40 36 Sample Location He/ He He/Ne He/ He He Total Ne N2 O2 Ar Ar Total Kr Ne/ Ar Ar/ Ar N2/ He/ (R/Ra) (He/Ne)Air (Rc/Ra) (ppm) (ppm) (%) (%) (%) (ppm) (ppm) Ar Ne MVT- El Tular 2.941 22.7 3.031 61.00 9.34 92.41 5.59 1.94 65.312 3.153 0.130 296 47.73 6.53 23 MVT- Cañada 2.044 2.8 2.621 3.23 4.00 80.78 17.77 1.38 47.017 2.772 0.077 294 58.40 0.81 24 Azul MVT- Agua 2.853 18.3 2.960 47.15 8.95 93.78 3.71 2.37 79.817 7.367 0.102 297 39.62 5.27 25 Blanca MVT- El 2.828 16.6 2.945 46.09 9.64 86.72 11.30 1.91 65.151 4.396 0.135 294 45.29 4.78 32 Zapote about 250 km east of Jungapeo. The plus symbols show et al., 1992). However, Jungapeo thermal waters have another group of meteoric waters collected from the lower 3H(≤1.8 TU, Table 4). The plot of tritium vs. Popocatépetl region (Werner et al., 1997; F. Goff, chloride (Fig. 17) shows that most thermal waters plot as unpublished data) that are matched by the range a group with no apparent trend. Thus, the aquifer presented in Inguaggiato et al. (2005). Los Azufres feeding most of the springs has multiple recharge paths meteoric waters and Jungapeo waters of all types are with different residence times inside the Tuxpan shield. isotopically similar because of geographic proximity El Aguacate, along the Río Tuxpan, stands out on and roughly equivalent elevations, whereas Jungapeo Fig. 17 because of its much higher chloride content, but waters are isotopically less depleted than most Mexico the relatively high 3H of this sample implies dilution City Basin waters and Popocatépetl waters because the with river water (Fig. 11D). The ages of the Jungapeo latter fluids issue farther from ocean sources of rain and mineral waters can be estimated using tables and at higher elevations. methods described in Goff et al. (1991) and Shevenell The shaded circle on Fig. 16 shows the pre- and Goff (1995), respectively for piston-flow and well- production isotopic composition of the ca. 300 °C Los mixed aquifers. Most Jungapeo thermal waters have Azufres geothermal reservoir (Barragán et al., 2005). high flow rates and are probably best modeled as piston- Los Azufres geothermal fluids are isotopically enriched flow aquifers. Because they have 0.3 to 1.8 TU 3H, their by about 6‰ ∂18O relative to Los Azufres meteoric waters have calculated mean residence times (water waters, a magnitude of enrichment commonly seen in ages) of about 7 to 25 years. high-temperature geothermal systems (Goff and Janik, 2000). Because Jungapeo waters show little, if any, 7. Gas geochemistry enrichment in oxygen-18 relative to the WMWL, there is no evidence of high-temperature rock–water isotopic Bulk chemistry of several Jungapeo area gas samples exchange in the Jungapeo system. (Table 7) show the gases are extremely rich in CO2 and Several of the water samples are slightly depleted in are usually contaminated with air components (N2,O2, oxygen-18 relative to the WMWL and the MBWL. and Ar). From a geothermal perspective, the gases have Although this apparent depletion may be mostly no detectable H2SorH2 and very low contents of CH4 analytical error, it may be that minor oxygen isotope and NH3. Gas from El Aguacate spring is not exchange is occurring between CO2 and H2O in the distinguishable from gases at other Jungapeo sites. tepid CO2-rich waters. Such an exchange is promoted by Carbon-13 isotope analyses of the CO2 from many sites cool temperatures and high CO2 gas flux, and has been (Table 4) show a narrow range of values between −6.7‰ documented in waters at other CO2-rich sites (Fritz and and −7.2‰. Although the carbon isotopes of Jungapeo Frape, 1982; Vuataz and Goff, 1986; Matthews et al., CO2 resemble the value for CO2 in air (about −7‰), the 1987; Pauwels et al., 1997). concentration of CO2 in air (0.04 mol%) is insignificant compared to the amount of CO2 in Jungapeo gases 6.3. Relative ages of waters (≥90 mol%). Thus air is not a viable source for most of the CO2. The isotopic composition of Jungapeo CO2 The cold background sample from El Concreto has indicates that most CO2 could logically originate from 2.1 TU tritium (3H), a relatively typical value for recent the mantle, assuming a MORB value of −5‰ to −8‰ 13 rain in southern Mexico and Central America (Janik ∂ C–CO2 (Hoefs, 1973; Rollinson, 1993). Because 26 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 temperatures of 150 °C exist at relatively shallow depths beneath the TMVB, some CO2 could be contributed to Jungapeo gases from thermal degradation of organic material in underlying Cretaceous marls and shales, a process that appears to produce most of the isotopically depleted CO2 in The Geysers, California (Bergfeld et al., 2001). Thermal decarbonation of Cretaceous limestone at moderate temperatures can not be a major source of CO2 in Jungapeo gases because the source carbonate would have to be more enriched in ∂13C than the produced CO2 and most marine limestones have values between 0±4‰ ∂13C(Hoefs, 1973; Goff et al., 1985). Noble gas data for four samples are listed in Table 8. The noble gases were analyzed from splits of headspace gas taken from the bulk gas samples. Unfortunately, a good noble gas split was not obtained from the El Aguacate site. One of the noble gas samples (Cañada Azul) is substantially contaminated with air. The other three samples contain less air contamination and are considered more representative of Jungapeo noble gases. The helium contents of the gases are low ≤ 3/4 ( 61 ppm) but the He ratios corrected for air 3/4 13 Fig. 19. Plot of He versus ∂ C–CO2 for gases of the Jungapeo area, contamination range from about 2 to 3 Rc/Ra. These Mexico, compared with gases from hot spots, MORB and various results indicate that a small mantle/magmatic compo- geothermal systems. L=Loihi volcano, Hawaii (Sedwick et al., 1994); nent of helium-3 is present in Jungapeo gases. SN=Sierra Negra and K=Kilauea (both from Goff et al., 2000); The relative concentration of helium in Jungapeo R=Reunion (Marty et al., 1993); MORB=mid-ocean ridge basalt gases can be examined in the N –Ar–He ternary of (Rollinson, 1993); VC =Valles caldera (Goff and Janik, 2002); 2 TG=The Geysers (Bergfeld et al., 2001); Popo=Popocatépetl Fig. 18. Jungapeo gases have no helium excess and plot (Inguaggiato et al., 2005); Tecuamburro from Janik et al. (1992);El on the join between air saturated water and air. In Olivar from Goff et al. (1987) and Kennedy et al. (1991). contrast, gases from high-temperature geothermal

systems (i.e., The Geysers, Yellowstone, and Valles caldera) or hot spot volcanoes (i.e. Kilauea and Sierra Negra) generally have relatively high bulk helium. 3/4 13 A comparison of He ratios and ∂ C–CO2 values between Jungapeo gases and other geothermal system and hot spot gases is shown in Fig. 19. Hot spot gases have He3/4 ratios varying from 13 to 28 Rc/Ra, whereas MORB and high-temperature geothermal gases have ratios between 6 and 9 Rc/Ra. By comparison, Jungapeo He3/4 ratios are small, only 2 to 3 Rc/Ra, about the same or slightly higher than those from Popocatépetl volcano area. Carbon-13 in CO2 measured in hot spot gases range narrowly between −2‰ and −6‰ whereas MORB values range between − 5‰ and − 8‰. Geothermal gases have values ranging from −2 to less − ‰ ∂13 – Fig. 18. Triangular plot of N2/100-Ar-10He for gases of the Jungapeo than 14 C CO2. Geothermal gases with very area, Mexico, compared with gases from selected hot spots and high- 13 depleted ∂ C–CO2 such as The Geysers contain temperature geothermal systems (modified from Giggenbach, 1992). substantial amounts of organic CO from marine Data sources: K (average Kilauea volcano, Hawaii) and SN (average 2 Sierra Negra volcano, Galapagos) from Goff et al. (2000); Valles sedimentary rocks (Bergfeld et al., 2001) as mentioned caldera and Yellowstone from Goff and Janik (2002); The Geysers above, whereas gases from geothermal systems under- from Lowenstern et al. (1999). lain by marine limestone (e.g., Valles caldera, C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 27

Tecuamburro, and El Olivar) often contain a mixture of chemistry of high-temperature fluids (Fournier, 1981; mantle and marine limestone sources (Janik et al., 1992; Goff and Janik, 2000). In spite of these difficulties, our Goff and Janik, 2002). Because Jungapeo is underlain in preferred temperature estimates are shown in bold on part by Cretaceous marls and shales, some CO2 gas Table 9. We have divided the waters into four groups: A could be derived from organic sources, although most deep water represented by El Aguacate, two types of CO2 seems to originate from a MORB-like source. very similar shallow waters exemplified by Agua Blanca and Cañada Azul, respectively, and background 8. Chemical geothermometry water represented by El Concreto. El Aguacate spring produces the only water that Thermal equilibration of the Jungapeo waters can be indicates derivation from a moderate- to high-temper- evaluated in the Na–K–Mg ternary (Fig. 20) devised by ature geothermal aquifer. The best estimated tempera- Giggenbach (1988). Most Jungapeo waters plot near the ture for this source fluid is around 150 to 175 °C (Fig. 20 Mg1/2 apex and are considered to be cold and and Table 9). The spring discharges along the Río “immature”. In contrast, EA waters form a linear cluster Tuxpan; thus it is undoubtedly mixed with some river straddling the 175 °C temperature line for full water. Discharge temperature–chloride relations also equilibration. The position of the cluster suggests that indicate mixing, as do many of the other plots described EAwaters are now partly equilibrated or mixed fluids, in above. Because the quartz conductive geothermometer which cooler Mg-rich waters dilute a relatively high- indicates a substantially lower temperature (128 °C) temperature fluid component. than 150 to 175 °C, mixing of high- and low- Calculated subsurface reservoir temperatures of temperature components is indicated. Temperature re- various Jungapeo fluids can be compared in Table 9 equilibration is implied by the Mg–Li and K–Mg using a standard suite of chemical geothermometers geothermometers, which indicate subsurface tempera- (Urbani, 1986). The calculations vary tremendously ture of only 80 to 100 °C (Giggenbach, 1988; Kharaka because Jungapeo waters show multiple processes such and Mariner, 1989). as equilibration at high and low temperatures, mixing of Jungapeo shallow mineral waters are derived from a fluids and re-equilibration of mixed fluids. Interpreta- large aquifer system that circulates in the young Tuxpan tion of chemical geothermometers is complicated in shield volcano E of the Río Tuxpan. Temperatures low- to moderate-temperature situations because many calculated from the amorphous silica and Na–K–Ca of the geothermometers are empirically derived from the (Mg-corrected) geothermometers are nearly identical to

Fig. 20. Triangular plot of Na/1000-K100-Mg1/2 for waters of the Jungapeo area, Mexico, compared with waters from high-temperature geothermal systems and from the low-temperature waters at Popocatépetl (Werner et al., 1997, Table 10 and Goff and Janik, 2000, Table 3). Labels: A=Ahauchapán; W=Wairakei, CP=Cerro Prieto; DV=Dixie Valley; PL=Platanares; V=Valles caldera; M=Miravalles; Z=Zunil; AH=Agua Hedionda; OX=Oaxtepec; IX=Ixtatlala. 28

Table 9 Estimated reservoir temperature of Jungapeo thermal waters using a standard suite of chemical geothermometers (Urbani, 1986). All values are in °C except where noted. Our preferred temperature estimate for each water type is shown in bold characters. Temperatures in parentheses violate the application rules of the geothermometer in question but are shown to complete the table a b b c d e e f g h i Site N Measured T(Amorph) T(Chal) T(Qtz T(Na/K)t T(Na/Li)d T(Na–K–Ca) T(Na–K–Ca) T(Na–K–Ca– R T(Mg/Li) T(K/Mg) T(D–P)

b f 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C. temperature cond) (B=1/3) (B=4/3) Mg) “Deep” water El Aguacate 5 33.4±1.2 (4) 9±7 100±9 128±8 104±11 132 (1) 150±8 {147±19} 37±13 38±7 98 (1) 83±5 49 (1)

“Shallow” water, Group I San José Purúa 12 32.0±0.3 (7) 32±9 129±11 154±10 191±22 198±2 (2) {172±9} 100±7 32±6(3) 51±3 72±1 (2) 64±4 47±3 (2) Agua Blanca 14 29.9±0.5 (9) 25±5 120±6 146±5 195±22 195±4 (2) {172±8} 95±5 35±5(5) 50±3 68±1 (2) 62±2 41±1 (2) El Puente 4 29.1±0.1 24±6 (3) 112±16 139±15 208±25 nd {176±11} 96±7 33 (1) 51±3 nd 63±4 nd Agua Amarilla 13 30.8±0.2 (8) 26±8 122±10 147±9 186±18 197 (1) {169±9} 95±8 34±4(3) 51±3 68 (1) 62±4 50 (1) El Tular 12 29.4±0.5 (7) 24±11 119±14 145±13 196±19 195 (1) {172±8} 93±6 31±3(7) 50±3 66 (1) 62±3 40 (1)

“Shallow” water, Group II Cañada Azul 13 27.1±0.9 (8) 20±9 114±11 141±10 214±41 193 (1) {175±13} 86±7 37±4(5) 50±3 61 (1) 59±3 42 (1) Varelas 3 29.8±0.3 20±8 115±9 141±9 185±5 174 (1) {162±5} 77±10 b50 (cool) 43±16 53 (1) 57±3 nd El Sauz 3 28.4±0.4 20±4 114±6 141±5 201±16 174 (1) {168±8} 79±8 b50 (cool) 56±1 48 (1) 53±5 nd El Gauyabo 2 29.7±0.3 16±13 109±16 136±15 173±26 179 (1) {160±7} 85±4 b50 (cool) 56±1 52 (1) 54±1 50 (1) El Zapote 3 29.7±0.6 21±9 116±11 142±10 198±1 180 (1) {170±2} 85±3 b50 (cool) 58±2 53 (1) 55±1 38 (1)

Background water El Concreto 1 24.5 19 113 140 229 165 {175} 74 b50 (cool) 64 36 47 nd a Number of analyses available for the averaged calculations; numbers in parentheses show the value of n in cases where parameters are lacking. b Amorphous silica, chalcedony, and quartz conductive equations in Fournier (1981). c Na–K equation of Truesdell as reported in Fournier (1981). d Na–Li equation for dilute waters of Fouillac and Michard (1981). e Na–K–Ca equations of Fournier and Truesdell (1973). f Na–K–Ca–Mg equation and calculated R-value of Fournier and Potter (1979). – 33 g Mg–Li equation of Kharaka and Mariner (1989). h K–Mg equation of Giggenbach et al. (1983) and Giggenbach (1988). i CO2–H2S–H2–CH4 equation of D'Amore and Panichi (1980). Average equilibrium temperature (if values are below detection limits, we used 0.001 H2, 0.0001 CH4, and 0.001 H2Sin geothermometer calculations). .Seee l ora fVlaooyadGohra eerh13(07 1 (2007) 163 Research Geothermal and Volcanology of Journal / al. et Siebe C. – 33

Fig. 21. NE–SW cross-section through the Jungapeo area, Mexico showing our conceptual model of the geothermal system. Profile shown is located on Fig. 2. 29 30 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 discharge temperatures (roughly 30 °C). Although some of organic remains in underlying marine rocks. trace elements show relatively high values (i.e., B), As Assuming a reasonable conductive thermal gradient of contents are extremely low and other constituents such 60 °C/km for this region (Prol-Ledsema and Juárez, as Ca, Mg, and Fe are much higher than those found in 1986; Prol-Ledesma, 1991) and an ambient temperature typical high-temperature geothermal fluids (Goff and of 18 °C (Fig. 10), the depth to a 175 °C aquifer as Janik, 2000). Paces (1975) noted that the Na–K–Ca indicated by the geothermometers is roughly 2.5 km. We geothermometer was not reliable for CO2-rich waters hesitate to call this deep aquifer a geothermal “reservoir” equilibrated below 75 °C and other geothermal because only one of the Jungapeo sites (El Aguacate) investigators have noted that Mg-rich waters cannot discharges water with high-temperature characteristics. equilibrate at temperatures above 75 °C (i.e., Fournier Meteoric waters mix with rising, gas-rich thermal and Potter, 1979). Based on such considerations, it is our fluids to form slightly acidic, bicarbonated waters that opinion that high CO2 and resulting HCO3 aggressively react with host rocks releasing soluble constituents. attack the basaltic andesite (especially olivine and Horizontal flow and permeability discordances along clinopyroxene crystals) in the Tuxpan shield volcano, faults force these waters to return into the permeable dissolving copious amounts of SiO2, Fe, Ca, Mg, and Tuxpan shield lava flows at more distal and lower some trace elements to produce mineralized waters elevations, where they corrode olivine and pyroxene superficially resembling higher temperature geothermal crystals (iddingsitization) and hydrate and leach the fluids. Thus, most geothermometers predict excessively matrix glass. The CO2-saturated groundwater finally high temperatures for the shallow Jungapeo waters. emerges deep in the Tuxpan River canyon at the contact Gas geothermometers indicate relatively cool tem- of the lavas with the underlying less permeable marine peratures of subsurface equilibration (≤50 °C). There basement. When cooling down to ambient temperatures are no apparent differences between the gases derived in pools and streamlets, opaline silica, calcite, goethite, from EA and other Jungapeo sites. etc., precipitate and form travertine coatings. The background water, El Concreto, contains some shallow mineralized water as described above. We were 10. Concluding remarks not able to find a completely uncontaminated sample of cold background water in the immediate vicinity of the The CO2-rich mineral waters of the Jungapeo area Tuxpan shield volcano. hint at first sight toward the existence of a high-enthalpy geothermal system. This idea is favored by the nearby 9. Geothermal system model occurrence of the Los Azufres geothermal field. The youngest (b0.60 Ma) volcanic edifices are monogenetic The main features of our conceptual model of the scoria cones, domes, and related lava flows whose geothermal system feeding the mineral springs are compositions point toward a deep magmatic source in shown in Fig. 21. In this model, NNW–SSE oriented the lithospheric mantle. Although closely linked in older reverse faults (Pasquaré et al., 1988) are super- space and time, whole-rock chemical analyses of the imposed by a set of younger parallel extensional normal scoria cones and lavas exclude the existence of a faults that produce a step-faulted pattern. These faults common large magma chamber at shallow crustal levels have served as conduits for ascending batches of magma that would be capable of inducing a high-enthalpy that upon eruption constructed the diverse volcanoes geothermal system. occurring in the area. These include the NNW–SSE In consequence, the Jungapeo geothermal system is oriented cones from which the voluminous lavas of the not capable of electricity production. Furthermore, the Tuxpan shield emanated. Accordingly, segments of bicarbonated waters might be difficult to use in space larger faults are occupied by cooling and degassing heating and other industrial applications due to the risk dikes that must reach deep into the crust and mantle and of CO2 release and precipitation of carbonates. Howev- serve as pathways for gases (e.g. He and some CO2)of er, the spring waters are ideal for balneological mantle origin. Meteoric water precipitating on the purposes. The benign properties of the waters to humans surface of the Tuxpan shield easily percolates downward are underscored by the fact that our analyses did not until reaching impermeable layers (marls and shales) of detect any toxic elements (e.g. high arsenic contents) or the local Cretaceous basement. This water is heated by an elevated level of radioactivity. In this sense we hope remnant heat from the magma bodies in the TMVB and that the municipal authorities of Jungapeo preserve the possibly by small shallow magma bodies of the basaltic natural beauty of this area for the enjoyment of future shield. Some CO2 may originate by thermal degradation generations. C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 31

Acknowledgments Chipera, S.J., Bish, D.L., 2002. FULLPAT: a full-pattern quantitative analysis program for X-ray powder diffraction using measured and calculated patterns. Journal of Applied Crystallography 35, C. Siebe was funded by grants CONACYT-50677-F 744–749. and DGAPA-UNAM-IN-101006-3. F. Goff, D. Counce, Choi, H.-S., Koh, Y.-K., Bae, D.-S., Park, S.-S., Hutcheon, I., Yun, S.-T.,

R. Poreda, and S. Chipera were funded by an 2005. Estimation of deep-reservoir temperature of CO2-rich springs Institutional Supporting Research and Development in Kangwon district, South Korea. Journal of Volcanology and – grant (Magmatic Tritium) to F. Goff from Los Alamos Geothermal Research 77, 77 89. – Cortés, A., Durazo, J., Farvolden, R.N., 1997. Studies of isotopic National Laboratory (1995 1997). A. Aguayo, N. hydrology of the Basin of Mexico and vicinity: annotated Cenizeros, and O. Cruz (UNAM) carried out chemical bibliography and interpretation. Journal of Hydrology 198, analyses of part of the water samples. Michael Abrams 346–376. (Jet Propulsion Laboratory) provided Landsat satellite Craig, H., 1961. Isotopic variations in meteoric waters. Science 133, – imagery funded by NASA. Climate data obtained at 1702 1703. D'Amore, F., Panichi, C., 1980. Evaluation of deep temperatures of Laguna El Freso meteorological station was kindly hydrothermal systems by a new gas geothermometer. Geochimica provided by Ing. Javier Espinosa Cruickshank (Servicio Cosmochimica Acta 44, 549–556. Meteorológico Nacional). Figures were drafted by Demant, A., Mavois, R., Silva Mora, L., 1975. Estudio geológico de Deborah Bergfeld (US Geological Survey) and by las hojas Morelia y Maravatío, Estado de Michoacán. Comisión Ignacio Hernández Javier, Lilia Arana Salinas, and Federal de Electricidad, Instituto de Geología, Universidad Nacional Autónoma de México. 53 pp. Renato Castro Govea (UNAM). Petrographic work was Dobson, P.F., Mahood, G.A., 1985. Volcanic stratigraphy of the Los carried out with the aid of a new Leitz polarizing Azufres geothermal area, Mexico. Journal of Volcanology and microscope donated by the Alexander von Humboldt Geothermal Research 25, 273–287. Foundation (Germany) to C. Siebe. Dawnika Blatter and Fahlquist, L., Janik, C.J., 1992. Procedures for collecting and Alain Demant freely shared information obtained during analyzing gas samples from geothermal systems. U.S. Geological Survey, Open-file Report 92-211. 19 pp. their earlier work in the study area. The manuscript was Fouillac, C., Michard, G., 1981. Sodium/lithium ratio in water applied kindly reviewed by S. Inguaggiato (Palermo, Italy), G. to geothermometry of geothermal reservoirs. Geothermics 10, Chiodini (Naples, Italy), and an anonymous reviewer. 55–70. Fournier, R.O., 1981. Application of water chemistry to geothermal exploration and reservoir engineering. In: Rybach, L., Muffler, L.J.P. References (Eds.), Geothermal Systems: Principles and Case Histories. Wiley, New York, pp. 109–143. Armienta, M.A., De la Cruz-Reyna, S., 1995. Some hydro- Fournier, R.O., Potter, R.W., 1979. Magnesium correction to the Na– geochemical fluctuations observed in Mexico related to volcanic K–Ca chemical geothermometer. Geochimica Cosmochimica Acta activity. Applied Geochemistry 10, 215–227. 43–9, 1543–1550. Baker, I., Haggerty, E.S., 1967. The alteration of olivine in basaltic and Fournier, R.O., Truesdell, A.H., 1973. An empirical Na–K–Ca associated lavas: Part II. Intermediate and low temperature alteration. geothermometer for natural waters. Geochimica Cosmochimica Contributions to Mineralogy and Petrology 16, 258–273. Acta 37, 1255–1275. Barnes, I., O'Neil, J.R., Rapp, J.B., White, D.E., 1973. Silica– Fritz, P., Frape, S.K., 1982. Comments on the 18O, 2H and chemical carbonate alteration of serpentine: wall rock alteration in mercury composition of saline groundwaters in the Canadian Shield. In: deposits of the California Coast Ranges. Economic Geology 68, Perry, E.C. (Ed.), Isotopic Studies of Hydrologic Processes. 388–398. Northern Illinois University Press, De Kalb, Illinois, pp. 57–63. Barragán, R.M., Arellano, V.M., Portugal, E., Sandoval, F., 2005. Giggenbach, W.F., 1988. Geothermal solute equilibria: Derivation of Isotopic (∂18O, D) patterns in Los Azufres (Mexico) geothermal Na–K–Ca–Mg geoindicators. Geochimica Cosmochimica Acta fluids related to reservoir exploitation. Geothermics 34, 527–547. 52, 2749–2765.

Bergfeld, D., Goff, F., Janik, C.J., 2001. Carbon isotopes and CO2 Giggenbach, W.F., 1992. Chemical techniques in geothermal explo- sources in The Geysers–Clear Lake region, northern California. ration. In: D'Amore, F. (Ed.), Applications of Geochemistry in Geothermics 30, 303–331. Geothermal Reservoir Development. Series of Technical Guides Blatter, D.L., Carmichael, I.S.E., 1998. Plagioclase-free andesites from on the Use of Geothermal Energy, UNITAR/UNDP Centre on Zitácuaro (Michoacán), Mexico; petrology and experimental Small Energy Resources, Rome, Italy, pp. 119–144. constraints. Contributions to Mineralogy and Petrology 132, Giggenbach, W.F., Gonfiantini, R., Jangi, B.L., Truesdell, A.H., 1983. 121–138. Isotopic and chemical composition of Parbati Valley geothermal Blatter, D.L., Carmichael, I.S.E., Deino, A.L., Renne, P.R., 2001. discharges, N.W. Himalayas, India. Geothermics 12, 199–222. Neogene volcanism at the front of the central Mexican volcanic Goff, F., unpublished. Geochemical data from Mexico volcanic and belt: basaltic andesites to dacites, with contemporaneous shosho- geothermal areas collected 1994 to 2002, Author files.

nites and high-TiO2 lava. Geological Society of America Bulletin Goff, F., Janik, C.J., 2000. Geothermal systems. In: Sigurdsson, H., 113, 1324–1342. Houghton, B., McNutt, S., Rymer, H., Stix, J. (Eds.), Encyclopedia Capra, L., Macías, J.L., Garduño, V.H., 1997. The Zitácuaro Volcanic of Volcanoes. Academic Press, San Diego, CA, pp. 817–834. Complex, Michoacán, Mexico: magmatic and eruptive history of a Goff, F., Janik, C.J., 2002. Gas geochemistry of the Valles caldera resurgent caldera. Geofísica Internacional 36, 161–179. region, New Mexico, and comparisons with gases at Yellowstone, 32 C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33

Long Valley and other geothermal systems. Journal of Volcanology MacDonald, G.A., Katsura, T., 1964. Chemical composition of and Geothermal Research 116, 299–323. Hawaiian lavas. Journal of Petrology 5, 82–133. Goff, F., McMurtry, G.M., 2000. Tritium and stable isotopes of Marques, J.M., Matias, M.J., Basto, M.J., Graça, R.C., Aires-Barros, magmatic waters. Journal of Volcanology and Geothermal L., Andrade, M., Carreira, P.M., Goff, F., Rocha, L., 2004. Water– Research 97, 347–396. rock interaction responsible for the origin of high pH mineral Goff, F., Gardner, J., Vidale, R., Charles, R., 1985. Geochemistry and waters (S-Portugal). Transactions Water–Rock Interaction 11, isotopes of fluids from Sulphur Springs, Valles Caldera, New Wanty & Seal (eds), p. 293–297. Mexico. Journal of Volcanology and Geothermal Research 23, Marty, B., Meynier, V., Nicolini, E., Griesshaber, E., Toutain, J.P., 273–297. 1993. Geochemistry of gas emanations: a case study of the Goff, F., Truesdell, A.H., Grigsby, C.O., Janik, C.J., Shevenell, L., Réunion Hot Spot, Indian Ocean. Applied Geochemistry 8, Paredes, R., Gutierrez, J.W., Trujillo, P.E., Counce, D., 1987. 141–152. Hydrogeochemical investigation of six geothermal sites, Hon- Matthews, A., Fouillac, C., Hill, R., O'Nions, R.K., Oxburgh, E.R., duras, Central America. Los Alamos National Laboratory, Report 1987. Mantle-derived volatiles in continental crust: the Massif LA-10785-MS. 200 pp. (obtainable from the second author). Central of France. Earth and Planetary Science Letters 85, 177–128. Goff, F., Goff, S.J., Kelkar, S., Shevenell, L., Truesdell, A., Musgrave, McLeod, R.G., 1989. Geology and tectonics of El Bosque area, J., Rufenacht, H., Flores, W., 1991. Exploration drilling, reservoir Michoacán, Mexico. Master's thesis, Brigham Young University, testing, age, and reservoir model of the Platanares geothermal Provo, Utah, 39 p. system, Honduras, Central America. Journal of Volcanology and Ostlund, H.G., Dorsey, H.G., 1977. Rapid electrolytic enrichment and Geothermal Research 45, 101–124. hydrogen gas proportional counting of tritium, in Low Radioac- Goff, F., McMurtry, G.M., Counce, D., Stimac, J.A., Roldán-Manzo, tivity Measurements. Proceedings of the International Conference A., Hilton, D., 2000. Contrasting hydrothermal activity at Sierra on Low Radioactivity Measurements and Applications, 6–10 Negra and Alcedo volcanoes, Galapagos Archipelago, Ecuador. October 1975, The High Tatras, Czechoslovakia, Slovenske Bulletin of Volcanology 62, 34–52. Pedagogicke, Nakladantelístvo, Bratislava. 5 pp. Hasenaka, T., Carmichael, I.S.E., 1985. The cinder cones of Paces, T., 1975. A systematic deviation from Na-K-Ca geothermo- -4 Michoacán–Guanajuato, central Mexico: their age, volume and meter below 75 °C and 10 atm PCO2. Geochimica Cosmochimica distribution, and magma discharge rate. Journal of Volcanology Acta 39, 541–545. and Geothermal Research 25, 105–124. Pasquaré, G., Garduño, V.H., Tibaldi, A., Ferrari, L., 1988. Stress Hasenaka, T., Carmichael, I.S.E., 1987. The cinder cones of pattern evolution in the central sector of the Mexican Volcanic Michoacán–Guanajuato, central Mexico; petrology and chemistry. Belt. Tectonophysics 146, 353–364. Journal of Petrology 28, 241–269. Pasquaré, G., Ferrari, L., Garduño, V.H., Tibaldi, A., Vezzoli, L., 1991. Hoefs, J., 1973. Stable Isotope Geochemistry. Springer, New York. Geologic map of the central sector of the Mexican Volcanic Belt, 140 pp. States of Guanajuato and Michoacán, Mexico. Geological Society Inguaggiato, S., Martin-Del Pozzo, A.L., Aguayo, A., Capasso, G., of America Map and Chart Series MCH072, pp. 1–19. Favara, R., 2005. Isotopic, chemical, and dissolved gas constraints Pauwels, H., Fouillac, C., Goff, F., Vuataz, F.-D., 1997. The isotopic

on spring water from Popocatépetl volcano (Mexico): evidence of and chemical composition of CO2-rich thermal waters in the Mont gas–water interaction between magmatic component and shallow Dore region (Massif-Central, France). Applied Geochemistry 12, fluids. Journal of Volcanology and Geothermal Research 141, 411–427. 91–108. Petersen, M.D., 1989. Geology and tectonics of the Jungapeo, Janik, C.J., Goff, F., Fahlquist, L., Adams, A., Roldán, A., Trujillo, P.E., Michoacán, Mexico area and its relationship to the Mexican Counce, D., 1992. Hydrogeochemical exploration of geothermal Volcanic Belt. Master's thesis, Brigham Young University, Provo, prospects in the Tecuamburro volcano region, Guatemala. Utah, 26 p. Geothermics 21, 447–481. Poreda, R., Craig, H., 1989. Helium isotope ratios in circum-Pacific Kennedy, B.M., Hiyagon, H., Reynolds, J.H., 1991. Noble gases from volcanic arcs. Nature 338, 473–478. Honduras geothermal sites. Journal of Volcanology and Geother- Pradal, E., Robin, C., 1994. Long-lived magmatic phases at Los mal Research 45, 29–39. Azufres volcanic center, Mexico. Journal of Volcanology and Kharaka, Y.K., Mariner, R.H., 1989. Chemical geothermometers and Geothermal Research 63, 201–215. their application to formation waters from sedimentary basins. In: Prol-Ledesma, R.M., 1991. Terrestrial heat flow in Mexico. In: Naeser, N.D., McCulloh, T.H. (Eds.), Thermal History of Cermak, V., Rybach, L. (Eds.), Exploration of the Deep Sedimentary Basins, Methods and Case Histories. Springer, New Continental Crust. Terrestrial Heat Flow and the Lithosphere York, pp. 99–177. Structure. Springer-Verlag, Berlin, pp. 475–485. Krauskopf, K.B., 1979. Introduction to Geochemistry. McGraw-Hill, Prol-Ledsema, R.M., Juárez, G., 1986. Geothermal map of Mexico. New York. 617 pp. Journal of Volcanology and Geothermal Research 28, 351–362. Les Bas, M.J., Le Maitre, R.W., Streckeisen, A., Zanettin, B., 1986. A Rollinson, H., 1993. Using Geochemical Data: Evaluation, Presenta- chemical classification of volcanic rocks based on the total alkali– tion, Interpretation. Wiley, New York. 325 pp. silica diagram. Journal of Petrology 27, 745–750. Schaaf, P., Stimac, J., Siebe, C., Macías, J.L., 2005. Geochemical Lowenstern, L.B., Janik, C.J., Fahlquist, L.S., Johnson, L.S., evidence for mantle origin and crustal processes in products from 1999. Gas and isotope geochemistry of 81 steam samples from Popocatépetl and surrounding monogenetic volcanoes, Central wells in The Geysers geothermal field, Sonoma and Lake Mexico. Journal of Petrology 46, 1243–1282. Counties, California, USA. U.S. Geological Survey, Open-file Sedwick, P.N., McMurtry, G.M., Hilton, D.R., Goff, F., 1994. Carbon Report 99-304. 28 pp. dioxide and helium in hydrothermal fluids from Loihi Seamount, Luhr, J.F., Simkin, T., 1993. Paricutín – The VolcanoBorn in a Mexican Hawaii: temporal variability and implications for the release of Cornfield. Geoscience Press, Inc., Phoenix, Arizona. 427 pp. mantle volatiles. Geochimica Cosmochimica Acta 58, 1219–1227. C. Siebe et al. / Journal of Volcanology and Geothermal Research 163 (2007) 1–33 33

Shevenell, L., Goff, F., 1995. The use of tritium in groundwater to Congress, vol. 2. International Geothermal Association, Inc., determine fluid mean residence times of Valles caldera hydrother- Auckland, New Zealand, pp. 1049–1052. mal fluids, New Mexico, USA. Journal of Volcanology and Urbani, F., 1986. GTHERM, a spreadsheet and graphic setup for Geothermal Research 67, 187–205. geothermal exploration. Unpub. Report, Los Alamos National Siebe, C., Rodríguez-Lara, V., Schaaf, P., Abrams, M., 2004. Laboratory, NM, 103 pp. Geochemistry, Sr–Nd isotope composition and tectonic setting of Vuataz, F.-D., Goff, F., 1986. Isotope geochemistry of thermal and Holocene Pelado, Guespalapa, and Chichinautzin scoria cones, non-thermal waters in the Valles Caldera, Jemez Mountains, south of Mexico City. Journal of Volcanology and Geothermal northern New Mexico. Journal of Geophysical Research 93, Research 130, 197–226. 6059–6067. Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics Werner, C., Janik, C., Goff, F., Counce, D., Johnson, L., Siebe, C., of oceanic basalts: implications for mantle composition and Delgado, H., Williams, S., Fischer, T., 1997. Geochemistry of processes. In: Saunders, A.D., Norry, M.J. (Eds.), Magmatism in summit fumarole vapors and flanking thermal/mineral waters at the Ocean Basins. Geological Society Special Publication, vol. 42, Popocatépetl Volcano, Mexico. Los Alamos National Laboratory, pp. 313–345. Report LA-13289-MS. 33 pp. (obtainable from the first or second Torres-Alvarado, I., Satir, M., Fortier, S., Metz, P., 1995. An oxygen authors). isotope study of hydrothermally altered rocks at the Los Azufres geothermal field, Mexico. Proceedings of the World Geothermal